ecologia mediterranea - UniversitÃ© d'Avignon et des Pays de Vaucluse

ecologia mediterranea 

SOMMAIRE – CONTENTS 

TOMÁS PAVLÍCEK, VLADIMIR CHIKATUNOV & EVIATAR NEVO 

Arthropods in the mounds of mole rats,Spalax ehrenbergi superspecies, 

in Israel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 

FAWZY M. SALAMA, MONIER M. ABD EL-GHANI, SALAH M. EL NAGGAR 

Tome 31 fascicule 1, 2005 

ISSN 0153-8756 

AND KHADIJA A. BAAYO 

Vegetation structure and environmental gradients in the Sallum area, Egypt . . . 15 

Tome 31 fascicule 1, 2005 

M. CHATZAKI, G. MARKAKIS & M. MYLONAS 

Phenological patterns of ground spiders (Araneae, Gnaphosidae) 

on Crete, Greece . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 

L.M.M. BIDAK 

On the seed ecology of two life forms of Spergularia (Caryophyllaceae) . . . . . . . 55 

SÉBASTIEN AUBRY & FRÉDÉRIC MAGNIN 

Factors structuring land snail communities in South-Eastern France: a comparison of 

two estimation methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 

CYRILLE B. K. RATHGEBER, ANTOINE NICAULT & JOËL GUIOT 

Évolution de la croissance radiale du pin d’Alep (Pinus halepensis Mill.) 

en Provence calcaire (sud-est de la France) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 

GÉRARD DE BÉLAIR 

Dynamique de la végétation de mares temporaires en Afrique du Nord (Numidie 

orientale, NE Algérie) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 

Faits de conservation en Méditerranée . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 

Analyses d’ouvrages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 

Résumés de thèse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 

Revue indexée dans Pascal-CNRS et Biosis 

ecologia mediterranea 

Tome 31 

Fascicule 1, 2005 

ISSN 0153-8756 

Revue internationale 

d’écologie méditerranéenne 

International Journal 

of Mediterranean Ecology

ecologia 

mediterranea 




of Mediterranean Ecology 

Tome 31 • Fascicule 1 • 2005

Rédacteur en chef • Managing editor Secrétariat • Secretariat 

FRÉDÉRIC MÉDAIL 

MICHELLE DOUGNY 

Rédacteurs • Editors 

LAURENCE AFFRE 

THIERRY DUTOIT 

JÉRÔME ORGEAS 

PHILIP ROCHE 

THIERRY TATONI 

ERIC VIDAL 

Fondateur • Founder 

PROFESSEUR PIERRE QUÉZEL 

Comité de lecture • Advisory board 

ARONSON J., CEFE-CNRS, Montpellier 

BARBERO M., IMEP, Université Aix-Marseille III 

BEAULIEU J.-L. DE, IMEP, Université Aix-Marseille III 

BROCK M., University of New England, Armidale, Australie 

CHEYLAN M., EPHE, Montpellier 

DEBUSSCHE M., CEFE-CNRS, Montpellier 

FADY B., INRA, Avignon 

GRILLAS P., Station biologique Tour du Valat, Arles 

GUIOT J., CEREGE-CNRS, Aix-en-Provence 

HOBBS R. J., CSIRO, Midland, Australie 

KREITER S., ENSA-M-INRA, Montpellier 

LE FLOC’H E., CEFE-CNRS, Montpellier 

MARGARIS N. S., University of the Aegean, Mytilène, Grèce 

OVALLE C., CSI-Quilamapu, INIA, Chili 

PEDROTTI F., Universita degli Studi, Camerino, Italie 

PLEGUEZUELOS J. M., Université de Grenade, Espagne 

PONEL P., IMEP, CNRS, Marseille 

PRODON R., EPHE, Montpellier 

RIDCHARSON D. M., University Cape Town, Afrique du Sud 

SANS F. X., Université de Barcelone, Espagne 

SHMIDA A., Hebrew University of Jérusalem, Israël 

TROUMBIS A., University of the Aegean Mytilene, Grèce 

URBINATI C., Agripolis, Legnaro, Italie 

Ecologia mediterranea 

Ecologia Mediterranea 

Europôle méditerranéen de l’Arbois, Bâtiment Villemin, B.P. 80 

F-13545 Aix-en-Provence, CEDEX 04, France 

Tél. : + 33 04 42 90 84 06 – Fax : + 33 04 91 28 80 51 

Internet : http://www.imep-cnrs.com/ecologia/ – email : f.medail@univ.u-3mrs.fr – 

Éditions Édisud / Atelier graphique Édisud 

3120 route d’Avignon, F-13090 Aix-en-Provence, France 

Tél. : 00 33 04 42 21 61 44 – Fax : 00 33 04 42 21 56 20 

Internet : http://www.edisud.com – email : info@edisud.com - 

© Édisud, 2006, tous droits réservés. 

Abonnements • Subscription (contacter Edisud) 

Un an : 2 numéros • One year : 2 issues 

— France : 61 € + 9,12 € de frais de port 

— Europe : 61 € + 12,2 € de frais de port 

— Amérique, Afrique, Asie : 61 € + 18,3 € de frais de port 

ISSN 0153-8756

ecologia 

mediterranea 




of Mediterranean Ecology 

Tome 31 • Fascicule 1 • 2005

Arthropods in the mounds of mole rats, 

Spalax ehrenbergi superspecies, in Israel 

Arthropodes dans les monticules de rats-taupes 

de la super-espèce Spalax ehrenbergi en Israël 

Tomás Pavlícek*, Vladimir Chikatunov** & Eviatar Nevo* 

* Institute of Evolution, University of Haifa, Mt. Carmel, Haifa 31905, Israel 

** Department of Zoology, Tel Aviv University, Ramat Aviv, Tel Aviv, Israel 

Address for correspondence: Dr. Tomás Pavlícek, Institute of Evolution, University of Haifa, 

Mt. Carmel, Haifa 31905, Israel. Tel. ++972-4-8240460, e-mail: patricia@research.haifa.ac.il. 

5 

Abstract 

A complex of invertebrates (mites, Collembola, earthworms, 

snails, Diplopoda, Myriapoda, Chilopoda, Isopoda and insects) 

was uncovered in large breeding mounds made by the blind subterranean 

mole rat (belonging to the species Spalax carmeli Nevo 

& Ivanitskaya & Beiles, 2000 of the Spalax ehrenbergi (Nehring, 

1898) superspecies) in the Haifa region, situated in the Northern 

Coastal Plain of Israel. The described properties of this complex 

were based on the 56 beetle species found there, mostly collected 

from the outer compact wall of the breeding mounds by forceps 

and exhausters. Most of the recorded beetle species were xerophiles 

or mesophiles and carnivores or detritivores. Only few species were 

phytophages, coprophages or necrophages. Since the invertebrate 

complex derives from the unique breeding mounds architecture, the 

mole rats can be considered as ecological architects of this microscale 

ecosystem. 

Key-words 

Spalax, breeding mounds, invertebrates, beetles, Israel 

Résumé 

Un complexe d’invertébrés (mites, collemboles, vers de terre, escargots, 

diplopodes, myriapodes, chilopodes, isopodes et insectes) a été 

mis au jour dans de grands monticules de reproduction construits 

par le rat-taupe souterrain aveugle (appartenant à l’espèce Spalax 

carmeli de la super-espèce Spalax ehrenbergi) dans la région de 

Haïfa, située dans la plaine côtière septentrionale d’Israël. Les 

caractéristiques de ce complexe ont été décrites à partir des 56 espèces 

de coléoptères trouvées à cet endroit, pour la plupart dans la 

paroi extérieure compacte des monticules d’élevage. La majorité des 

coléoptères recensés sont xérophiles ou mésophiles et carnivores ou 

détritivores. Seules quelques espèces sont phytophages, coprophages 

ou nécrophages. Étant donné que le complexe d’invertébrés est le 

produit de l’architecture unique en son genre des monticules de 

reproduction, les rats-taupes peuvent être considérés comme des 

architectes écologiques de ce micro-écosystème. 

Mots-clé 

Spalax, monticules de reproduction, invertébrés, coléoptères, Israël 

ecologia mediterranea, tome 31, fascicule 1, 2005, p. 5-13

◆ T. PAVLICEK, V. CHIKATUNOV & E. NEVO 

6 

INTRODUCTION 

Large soil structures such as breeding mounds of mole 

rats (fig. 1; Nevo, 1961, 1991, 1999) might be used as a 

shelter and as a breeding and living place by a number 

of invertebrates. To obtain preliminary knowledge about 

invertebrates living in the breeding mounds of mole rats, 

we studied such mounds near of the area known as Check 

Post in Haifa, Israel. These huge breeding mounds, built 

by Spalax carmeli Nevo & Ivanitskaya & Beiles, 2000 

(S. ehrenbergi Nehring, 1898 superspecies) might be 

the most elaborate in Israel, probably due to the poor 

soil drainage at the locality (Nevo, 1961). The mole rat 

superspecies, Spalax ehrenbergi, distributed in the Middle 

East (Nevo, 1991, 1999; Nevo et al., 2000), is a blind, 

solitary, subterranean mammal. Both, females and males, 

build relatively small nutritional mounds. However, both 

genders build also during the winter breeding season 

(October to March/April), particularly in poorly-drained 

soil, much bigger mounds that are either dome-shaped 

(females) or cone-shaped (males) (Nevo, 1961). During 

the breeding season, females use the big winter mounds 

for breeding and males for resting. After this period, 

during summer time, males and females return to deeper 

runways. The burrowing activity is then limited because 

of the soil dryness. 

Female breeding mounds, the subject of our study, 

constitute a system composed of a central breeding nest, 

storage chambers containing bulbs and rhizomes, and 

sanitary chambers (toilets), all connected by spiraling 

tunnels and roofed by a solid dome (fig. 1). Female 

breeding mounds are connected with the peripheral 

region by feeding tunnels (Nevo, 1961). In flooded 

areas with poorly drained dark alluvial clayey soils, the 

breeding mounds might be exceptionally large. They have 

well-plastered tunnels and are more solidly constructed 

than breeding mounds built in areas with good drainage 

(Nevo, 1961). 

Figure 1. Breeding mound of the blind mole rat, Spalax ehrenbergi. 

Cross section in a natural breeding mound (note the nest in the centre and the storage chambers in the periphery) 

(from Nevo, 1999, with permision of the author). 

ecologia mediterranea, tome 31, fascicule 1, 2005

ARTHROPODS IN THE MOUNDS OF MOLE RATS, SPALAX EHRENBERGI SUPERSPECIES, IN ISRAEL ◆ 

MATERIALS AND METHODS 

The study of the invertebrate complex in the breeding 

mounds of mole rat, Spalax carmeli (2n = 58) was 

conducted in an uncultivated field near of Check-Post, 

Haifa, Israel. Zoogeographically, the studied area is part 

of the Northern Coastal Plain (see zoogeographical areas 

in Israel in Fishelson, 1985). 

The studied field consists of black alluvial soil (Atlas 

of Israel, 1970), has poor drainage, and the water table 

is 0-50 cm below the surface (Nevo, 1961). The field 

is regularly flooded in January/February and surface 

waters may stay there for many days. To avoid damage 

to the breeding mounds by flooding, females build large 

breeding mounds averaging 175 cm in length, 140 cm in 

width and 40 cm in height (Nevo, 1961). These mounds 

are always projecting above the water level in order to 

avoid the suffocation of mother and pups. Bulbs of numerous 

geophytes (Muscari comosum L., Gladiolus italicus 

Mill. and Narcissus tazetta L.) growing in the field form 

the basic food of the mole rats living there. Note that 

the field chosen near Check-Post does not represent a 

solitary habitat with uniquely large breeding mounds in 

Israel. Nevo (1961) described similarly large and even 

larger breeding mounds in the other flooded areas in 

Israel as well. 

We collected invertebrates once per spring, in 

Februrary-March, from 1997 till 2000. In each collection 

session, 5 female breeding mounds were checked, and 

one quarter of the volume of every studied mound was 

searched for invertebrates, which were collected individually 

by means of forceps and exhausters. The beetles 

were identified by comparison with a reference collection 

kept in the Department of Zoology at the Tel Aviv 

University. The collected specimens of invertebrates, with 

the exception of beetles, are deposited in the Institute of 

Evolution at the University of Haifa, whereas the beetles 

are deposited in the insect collection at the Department 

of Zoology at the Tel Aviv University. 

We collected the following groups of invertebrates in 

the breeding mounds of female mole rats: earthworms, 

land-snails, Myriapoda (Scolopendra cingulata Latreille, 

1789), Diplopoda (Tetrarthrosoma syriacum Humbert & 

Saussure, 1869), Isopoda, Chilopoda, insects (Heteroptera, 

ants, solitary bees and beetles) and spiders. Additionally, 

mites and Collembola were observed. Since we are 

primarily familiar with beetles, most of our conclusions 

will be based on the analysis of this group, which was the 

richest in species numbers and the most abundant group 

in the breeding mounds. 

Beetle diversity 

Altogether, we collected 422 specimens (7.5 ± 13.4 

specimens per beetle species) representing 56 beetle species, 

45 beetle genera and 9 beetle families (appendix 1). 

Surprisingly, 18 of these species (32 %) had not been 

recorded earlier in the Northern Coastal Plain (appendix 

1) and two of them represented new records for Israel 

(Brachinus peregrinus Apfelbeck, 1904 and Harpalus basanicus 

Sahlberg, 1913). To our knowledge, the last record of 

the rare Pterostichus pertusus Schaum, 1858 in Israel was 

that of Bodenheimer (1937). Ground and darkling beetles 

(families Carabidae and Tenebrionidae) represented the 

largest beetle groups in the mounds (41 species, 73% of 

all species assemblage, fig. 2). Those 41 species represent 

about 31 % of the ground and darkling beetle species 

known from the Northern Coastal Plain (Chikatunov, 

unpublished). The rest of the beetle species belonged to 

the following beetle families: Staphylinidae (n = 5; 8.9 %), 

7 

RESULTS 

Figure 2. Proportions of species from different beetle families in the breeding 

mounds of the mole rat Spalax carmeli, Check Post, Haifa, Israel. 



8 

Curculionidae (n = 3; 5.4 %), Scarabaeidae (n = 3; 

5.4 %), Coccinellidae (n = 1; 1.8 %), Chrysomelidae 

(n = 1; 1.8 %), Elateridae (n = 1; 1.8 %) and Silphidae 

(n = 1; 1.8 %) (fig. 2). 

The outer and inner solid dome layers 

Beetles and other invertebrates were collected from 

the solid dome except for Aphodius spp., which were 

collected in the sanitary chambers. According to our 

observation, there might be at least two different ecological 

niches in the wall of the solid dome: the outer and 

inner wall layers. The outer wall layer of the solid dome 

is exposed directly to the surrounding environment and 

to direct solar radiation. Therefore, it might be microclimatically 

more fluctuating and generally drier and 

warmer than the inner wall layer. We observed that most 

of the collected specimens belonging to the beetle genera 

Belopus, Cabirutus, Cossyphus, Eutagenia, Gonocephalum 

and Coccinella were found on and in the outer wall. The 

inner wall layer of the solid dome borders the outer layer 

on one side and tunnels, the chambers and the breeding 

nest on the second side. It has no direct contact with the 

surrounding environment and therefore might be microclimatically 

more stable and wetter. Most of the specimens 

representing the remaining genera (N = 38, 84 % 

of all genera) not collected in the outer layer (appendix 1) 

were found in the inner layer. The only exception were 

Aphodius spp. collected in sanitary chambers. Apart from 

adults in the solid dome wall, we also found larvae and 

pupae of Carabus hemprichi Dejean, 1826 and nonidentified 

beetle larvae. This indicates that some species 

develop completely inside the breeding mounds. 

Trophic network 

As regards trophic relationships, we recorded phytophage 

(15 species, 26.8 %), carnivore (23 species, 

41.1 %), coprophage (3 species, 5.4%), necrophage 

(1 species, 1.8 %) and detritivore (14 species, 25 %) 

beetles in the breeding mounds (fig. 3, appendix 1). In 

seven phytophage species (12.5 % of all beetle species) 

both adults and larvae were phytophages and in eight 

species (14.3 % of all beetle species) adults were phytophages 

and their larvae carnivores. Phytophages were 

represented by 54 specimens (12.8 % of the total beetle 

abundance). The most frequent phytophage species was 

the leaf beetle Gonioctena fornicata Br ggemann, 1873. 

Carnivore species involved 234 specimens (55.5 % 

of the total beetle abundance). Sixteen carnivore species 

Figure 3. Proportions of beetle species in different trophic zones in breeding 

mounds of Spalax carmeli, Check Post, Haifa, Israel. 

were mesophiles (69.6 % of carnivore species) and all 

carnivore and mesophile species were members of the 

families Carabidae and Staphylinidae (appendix 1). 

The development of coprophage beetle species is connected 

with dung, and, indeed, they were found in the 

sanitary chambers of the breeding mounds, in contrast 

to all other invertebrates collected from the solid dome. 

Rather surprisingly, coprophages were found in very low 

numbers, and every coprophage species was represented 

by one specimen only. One necrophage species was represented 

by four specimens (1 % of all beetle specimens). 

The detritophage species involved 127 specimens 

(30.1 % of the total beetle abundance). All tenebrionid 

beetles were detritophages and xerophiles, which might 

explain why many of these beetles were represented in the 

outer layer of the soil dome. 

As regards humidity preference, 19 species (34 % of 

all beetle species) were xerophiles (fig. 4) represented 

by 173 specimens (41 % of the total beetle abundance). 

Twenty-four species (43% of all beetle species) were 

mesophiles (fig. 4) and involved 174 specimens (41 % 

of the total beetle abundance). Only four species (7 % 

of the beetle assemblage) were hydrophiles (fig. 4) involving 

58 specimens (14 % of the total beetle abundance). 

Water requirements were not known for nine species 



Figure 4. Proportions of xerophile, hydrophile and mesophile beetle species 

in breeding mounds of the mole rat, Spalax carmeli, Check Post, Israel. 

(16 %) represented by 17 specimens (4 % of the total 

beetle abundance). Both mesophile and xerophile species 

accounted for 77 % of the species richness and for 82 % 

of the beetle abundance. 

DISCUSSION 

The breeding mound, 

a unique microhabitat and ecosystem 

The mole rat breeding mound represents a unique 

microhabitat and ecosystem for invertebrates appearing 

in high density. No other study of invertebrates, apart 

of ectoparasites (Costa & Nevo, 1969) has previously 

been undertaken in the breeding mounds of mole rat, S. 

ehrenbergi. Even the number of recorded species could 

be probably higher if other collection techniques were 

employed and sampling covered a larger area. Similarly, 

up to now, little attention had been paid to the study of 

the invertebrate fauna of mounds and nests of other 

mole-rat species (A. Scharff recently investigated insect 

associated with burrows of mole rat species belonging to 

ecologia mediterranea, tome 31, fascicule 1, 2005, p. 5-13 

the genus Cryptomys, pers. com.), again apart of parasites 

(for example Gupta & Trivedi, 1985; Sayin et al., 1977; 

Scharff et al., 1996, 1997). It does appear, however, 

that breeding mounds and galleries (Cox, 1972, 1984, 

1990; Cox & Gakahu, 1984; Cox & Hunt, 1990; Cox & 

Scheffer, 1991; Nevo, 1999) provide a suitable structures 

for the evolution of rich underground micro-ecosystems 

generated by subterranean mammals. The studied breeding 

mounds were colonized by invertebrates with specific 

(presumably microclimatic) requirements, as might 

be deduced from the fact that most collected species were 

xerophiles or mesophiles and carnivores or detritivores. 

Remarkably, 18 of the species found in the breeding 

mounds constitute new records for the Northern Coastal 

Plain and two out of them even new records for Israel 

(appendix 1). 

Beetles and other invertebrates were collected only 

in the solid dome, with the exception of Aphodius sp. 

and A. granarius Linnaeus, 1767, which were collected 

solely in the sanitary chambers. Interestingly enough, 

Aphodius sp. represents a new species to science (David 

Král, pers. comm.) and might be exclusively associated 

with excrements of mole rats. The two other reported 

coprophages, A. granarius and Copris hispanus Linnaeus, 

1758, are also frequent in the dung of land mammals. 

Interestingly, phytophages were not found inside the storage 

chambers which contained large amounts of stored 

and kept bulbs and rhizomes. Nevertheless, we did not 

collect very small invertebrates such as nematodes and 

mites. Theirs and the other small invertebrates presence 

in the breeding mounds of the mole rats is corroborated 

by the fact that 5 species of fleas and 53 species of 

gamasid mites had been collected previously from mole 

rat breeding nests in Israel (Costa & Nevo, 1969) and 

additional new flea species in mole rat breeding nests 

in Turkey (Aktas, 1989). Besides the above mentioned 

fleas and gamasid mites, it might be expected that endoparasites 

at specific life stages are present in the sanitary 

room. Spalax ehrenbergi superspecies seems to host series 

of different endoparasites such as: helminthes and larval 

nematodes (Wertheim & Nevo, 1971; Fair et al., 1990), 

coccidia (Golemansky & Darwish, 1992, Couch et al., 

1993) and Eimerian protoctists (Sayin, 1980). 

The solid dome 

The most important part of the breeding mounds for 

invertebrates appears to be the solid dome where most of 

the species and specimens concentrate. The solid dome 

constitues many interstices, where invertebrates might 

9


10 

move and develop. Till now, a few vertebrates had been 

observed in mounds, such as the warm snake (Typhlops 

vermicularis Merrem, 1820: Nevo, Ivanitskaya, Pavlíček 

& Chikatunov, pers. observ.), the mouse Mus macedonicus 

Petrov & Ružic, 1983, and the lesser white-toothed shrew, 

Crocidura suaveolens Pallas, 1811 (Nevo, pers. observ.). 

According to our observation, at least two different ecological 

niches (layers) or species gradients colonized by 

different species can be distinguished in the solid dome, 

namely the outer and inner wall layer, colonized by different 

species. Nevertheless, further research will be 

needed to confirm our observation and to gather precise 

quantitative and qualitative data. The observed distribution 

of beetles in outer and inner side of the solid dome 

might be connected with their favorite temperature ranges 

for activity. According to Bodenheimer (1934), the temperature 

range for high activity of the darkling beetles 

(prevailing on the solid dome surface) is 30.1 0 C-41.3 0 C, 

whereas in the ground beetles (prevailing inside of the 

solid dome) it is only 31 0 C-35 0 C. The specific microclimate 

of breeding mounds might be the reason why 

xerophiles and mesophiles constituted almost 86 % of the 

beetle species assemblage. The interesting species from 

this group is Phaleria acuminata that is usually present on 

sea-shores (Canzoneri, 1968). In the Check-post area, it 

might be present since the distance from the sea shore is 

less than 1 km, or this species is also able to colonize some 

habitats more inland. The second option supports the 

fact that P. acuminata was collected also from the Upper 

Galilee (Chikatunov, unpublished) in a distance of tens 

kilometers from the coastal plain. Our observations suggest 

that the mole rat breeding mounds support mainly 

the activity of xerophile and mesophile beetles which are 

carnivores or detritivores. The carnivore trophic zone 

was richer in species number and abundance than the 

detritivore one. Carnivores might feed on detritivore 

species, on themselves, on species of other trophic zones 

and on many small invertebrates. However, the presence 

of some (e.g., Coccinella septempunctata Linnaeus, 

1758 of which adults and larvae are predators) might be 

accidental or they are hiding in mounds during cold and 

rainy period. Detritivore species are perhaps prosperous 

in the mounds due to the large amount of organic remains 

in the soil. A similar case was described from the nests 

of the harvester ant, Messor bouvieri Bondroit, 1918, in 

South Europe. Inside the nests, darkling beetles and other 

invertebrates were up to 660 time more concentrated than 

outside, probably due to the high concentration of organic 

remains gathered together by ants (Sanchezpinero & 

Gomez, 1995). In the Check Post area, a relatively rich 

trophic layer was that of phytophages. We found more 

different phytophage species than detritivore ones, but 

they were much less abundant than the latter. In part, 

they might develop in the solid dome. However, the 

presence of some phytophagous species (e.g., Kalcapion 

semivittatum Gyllenhal, 1839, Gonioctena fornicata) 

might be accidental or they are temporally hiding there. 

Otherwise, the presence of K. semivittatum might be 

difficult to explain because it lives and develops in galls on 

Mercurialis annua L. Nevertheless, the foodweb structure 

in the mole rat breeding mound seems to start with the 

phytophage and detritivore tropic zones followed by the 

carnivore trophic zone. 

CONCLUSIONS 

Mole rats are extraordinary ecological architects and 

subterranean rodent farmers. In heavy, badly drained 

alluvial soils they construct unique complex structures 

to increase their Darwinian fitness by escaping floods and 

avoiding mortality of the mother and offspring. These 

unique breeding mounds are the cradles of a remarkable 

concentration of primarily invertebrate fauna with predominating 

beetles – a phenomenon not studied earlier 

in the Spalax ehrenbergi superspecies. The beetles display 

a trophic network of phytophages, detritivores and carnivores. 

The warm nest in the breeding mound and the 

organic detritus harbour small arthropods (such as mites) 

providing food for the bigger detritivore and carnivore 

beetles and other arthropods (e.g., myriapods). Thus, the 

mole rat breeding mounds represent a construction which 

supports this unique microscale ecosystem. 

ACKNOWLEDGEMENTS 

We thank the Chair of Evolutionary Biology and the 

Ancell-Teicher Research Foundation for Genetics and 

Molecular Evolution for financial support. We are grateful 

also to Ms. Robin Permut (Institute of Evolution) for 

English corrections and to Ms. Patricia Cardet (Haifa) for 

the translation of the Title and Abstract into French. 



References 

AKTAS M., 1989. Ctenophthalmus harputus, a new Spalax flea 

from Turkey. Med. Vet. Entomol., 3 : 23-27. 

ATLAS OF ISRAEL, 1970. Jerusalem: Ministry of Labour and 

Elsevier, Amsterdam. 1 volume (various pagins). 

BODENHEIMER F.S., 1934. Studies on ecology of Palestinian 

Coleoptera: II. Seasonal and diurnal appearances and activity. 

Bull. Soc. Entomol. Egypte, 18: 211-241. 

BODENHEIMER F.S., 1937. Prodromus Faunae Palestinae. Essai sur 

des éléments zoogéographiques et historiques du sud-ouest 

du sous-règne Paléarctique. Mém. Inst. Égypte, 33 : 1-286. 

CANZONERI S., 1968. Materiali per una monographia delle 

Phaleria del sottogenere Phaleria Latr. XX Contributo alla 

conescenza dei Coleoptera Tenebrionidae. Mem. Soc. Entomol. 

Ital., 47: 117-167. 

COSTA M. & NEVO E., 1969. Nidicolous arthropods associated 

with different chromosomal types of Spalax ehrenbergi 

Nehring. Zool. J. Linn. Soc., 48: 199-215. 

COUCH L., DUSZYNSKI D.W. & NEVO E., 1993. Coccidia (Apicomplexa), 

genetic diversity, and environmental unpredictability 

of four chromosomal species of the subterranean superspecies 

Spalax ehrenbergi (mole rat) in Israel. Parasitology, 79: 

181-189. 

COX C.B., 1972. A new digging dicynodont from the Upper 

Permian of Tanzania. In: Josey K.A. & Kemp T.S. (eds.), 

Studies in Vertebrate Evolution. Winchester Press, New York: 

173-179. 

COX G.W., 1984. The distribution and origin of mima mound 

grasslands in San Diego, California. Ecology, 65: 1397-1405. 

COX G.W., 1990. Soil mining by pocket gopher along topographic 

gradients in a mima moundfield. Ecology, 71: 837-843. 

COX G.W. & GAKAHU C.G., 1984. The formation of mima 

mounds in the Kenya highlands. A test of the Dalquest- 

Scheffer hypothesis. J. Mammal., 65: 149-152. 

COX G.W. & HUNT J., 1990. Form of mima mounds in relation 

to occupancy by pocket gophers. J. Mammal., 71: 90-94. 

COX G.W. & SCHEFFER V.B., 1991. Pocket gophers and mima 

terrain in North America. Nat. Areas J., 11: 193-198. 

FAIR J.M., SCHMIDT G.D. & WERTHEIM G., 1990. New species of 

Andrya and Paranoplocephala (Cestoidea: Anoplocephalidae) 

from voles and mole rats in Israel and Syria. J. Parasitol., 76: 

641-644. 

FISHELSON L., 1985. Fauna Palaestina. Orthoptera: Acridoidea. 

The Israel Academy of Sciences and Humanities, Jerusalem, 

229 p. 

GUPTA S.P. & TRIVEDI K.K., 1985. Nematode parasites of 

vertebrates. A new nematode, Gongylonema fotedari sp. 

nov. (family: Spiruridae Oerley, 1985) from Indian mole 

rat Bandicota bengalensis from Lucknow, U.P. Indian J. 

Helminthol., 37: 100-108. 

GOLEMANSKY V.G. & DARWISH A.I., 1992. The coccidian 

parasites (Apicomplexa: Coccidia) of the herbivorous mole 

rat - Spalax (Microspalax) ehrenbergi Nehring (Rodentia: 

Spalacidae) from Syria. Acta Protozool., 31: 173-176. 

NEVO E., 1961. Observations on Israeli populations of the mole 

rat, Spalax ehrenbergi Nehring 1898. Mammalia, 25: 127- 

144. 

NEVO E., 1991. Evolutionary theory and processes of active 

speciation and adaptive radiation in subterranean mole rats, 

Spalax ehrenbergi in Israel. Evol. Biol., 25: 1-125. 

NEVO E., 1999. Mosaic evolution of subterranean mammals: regression, 

progression and global convergence. Oxford University 

Press, Oxford, 413 p. 

NEVO E., IVANITSKAYA E. & BEILES A., 2000. Adaptive radiation 

of blind subterranean mole rats: naming and revising the four 

sibling species of the Spalax ehrenbergi superspecies in Israel: 

Spalax galili (2n=52), S. golani (2n=54), S. carmeli (2n=58) 

and S. judaei (2n=60). Backhuys Publishers, Leiden, 198 p. 

SANCHEZPINERO F. & GOMEZ J.M., 1995. Use of ant-nest debris 

by darkling beetles and other arthropod species in an arid 

system in South Europe. J. Arid Environ., 31: 91-104. 

SAYIN F., 1980. Eimeriidae of the herbivorous mole rat, Spalax 

ehrenbergi Nehring. J. Protozool., 27: 364-367. 

SAYIN F., DINCER S. & MERIC I., 1977. Coccidia, Protozoa 

Eimeridae of the herbivorous mole-rat, Spalax leucodon. 

J. Protozool., 24: 210-212. 

SCHARFF A., BURDA H., TENORA F., KAWALIKA M. & BARUS V., 

1997. Parasites in social subterranean Zambian mole-rats 

(Cryptomys spp., Bathyergidae, Rodentia). J. Zool., 241: 

571-577. 

SCHARFF A., TENORA F., KAWALIKA M., BARUS V. & BURDA 

H., 1996. Helminths from Zambian mole-rats (Cryptomys, 

Bathyergidae, Rodentia). Helminthol., 33: 105-110. 

WERTHEIM G. & NEVO E., 1971. Helminths of birds and mammals 

from Israel: III. Helminths from chromosomal forms of 

the mole-rat, Spalax ehrenbergi. J. Helminthol., 45: 161-169. 

11 



APPENDIX 1 

List of beetles species recorded from the breeding mounds of Spalax carmeli, 

in the Check Post area, near Haifa, Israel 

(N = number of collected specimens, * new record for Israel, ** new record for the Northern Coastal Plain, 

M – mesophile(s), X – xerophile(s), H –hydrophile(s). 

12 

SPECIES N TROPHIC ZONE CLIMATIC CORRELATE 

Family Carabidae 

Amara aenea Degeer, 1774 1 Carnivores M 

Anchomenus dorsalis Pontoppidan, 1763 7 Carnivores M 

Acinopus laevigatus Ménétries, 1823 5 Adults phytophagous, larvae carnivorous X 

Bembidion quadrifossulatum 

coelesyria Netolitzky, 1921** 

2 Carnivores H 

Bembidion lampros Herbst, 1784 1 Carnivores H 

Brachinus berytensis Reiche, 1855 25 Carnivores M 

Brachinus exhalans hebraicus Reiche, 1855 1 Carnivores M 

Brachinus peregrinus Apfelbeck, 1904* 12 Carnivores M 

Carabus hemprichi Dejean, 1826** 8 Carnivores M 

Carterus rufipes Chaudoir, 1843** 5 Phytophages X 

Chlaenius aeneocephalus Dejean, 1826 54 Carnivores H 

Dyschirius beludscha ganglbaueri Znojko, 1927 1 Carnivores H 

Ditomus calydonius Rossi, 1790 1 Phytophages X 

Harpalus basanicus Sahlberg, 1913** 3 Adults phytophagous, larvae carnivorous M 

Harpalus smyrnensis Heiden, 1888** 14 Adults phytophagous, larvae carnivorous M 

Microlestes seladon Holdhaus, 1912** 1 Carnivores M 

Notiophilus danieli Reitter, 1897** 2 Carnivores M 

Ophonus battus Reitter, 1900 1 Adults phytophagous, larvae carnivorous M 

Ophonus cribrellus Reiche, 1855 1 Adults phytophagous, larvae carnivorous M 

Ophonus oblongus Schaum, 1858** 5 Adults phytophagous, larvae carnivorous M 

Orthomus berytensis Reiche et Saulcy, 1857 5 Adults phytophagous, larvae carnivorous M 

Parophonus fallax Peyron, 1858** 5 Adults phytophagous, larvae carnivorous M 

Pterostichus pertusus Schaum, 1858** 1 Carnivores M 

Scarites planus Bonelli, 1813 15 Carnivores X 

Scarites saxicola Bonelli, 1813** 20 Carnivores X 

Siagona europaea Dejean, 1816 2 Carnivores M 

Siagona longula Reiche et Saulcy, 1855 69 Carnivores M 

Family Chrysomelidae 

Gonioctena fornicata Br ggemann, 1873** 4 Phytophages ? 

Family Coccinellidae 

Coccinella septempunctata Linnaeus, 1758 5 Carnivores ? 

Family Curculionidae 

Brachycerus junix Lichtenstein 1796 1 Phytophages ? 

Hypera sp. 1 Phytophages ? 

Kalcapion semivittatum Gyllenhal, 1839 2 Phytophages ? 



Family Elateridae 

Cardiophorus discicollis Herbst, 1806** 1 Adults phytophagous, larvae rhizophagous ? 

Family Scarabaeidae 

Aphodius granarius Linnaeus, 1767 1 Coprophages ? 

Aphodius sp. 1 Coprophages ? 

Copris hispanus Linnaeus, 1758 1 Coprophages ? 

Family Silphidae 

Ablattaria arenaria Kraatz, 1876 4 Necrophages M 

Family Staphylinidae 

Leptolinus nothus Erichson, 1839** 1 Carnivores M 

Xantholinus graecus Kraatz, 1858 1 Carnivores M 

Platyprosopus hierichonticus Reiche et Saulcy, 1856 3 Carnivores M 

Philonthus corruscus Gravenhorst, 1802** 1 Carnivores M 

Pseudocypus excisus G. Müller, 1925** 1 Carnivores M 

Family Tenebrionidae 

Adelostoma sulcatum cordatum Solier, 1837 1 Detritophages X 

Belopus syriacus Zoufal, 1893 11 Detritophages X 

Cabirutus castaneus Reitter, 1904 3 Detritophages X 

Clitobius variolatus Allard, 1884 9 Detritophages X 

Cossyphus rugulosus Peyron, 1854 48 Detritophages X 

Crypticus gibbulus Quensel, 1806 2 Detritophages X 

Dichillus nitidulus Reitter, 1886 1 Detritophages X 

Eutagenia cribricollis Reitter, 1886 1 Detritophages X 

Gonocephalum costatum rugulosum Küster, 1849 28 Detritophages X 

Gonocephalum rusticum Olivier, 1811 4 Detritophages X 

Mesomorphus longulus Reiche et Saulcy, 1857** 12 Detritophages X 

Phaleria acuminata Küster, 1852 1 Detritophages X 

Stenosis comata Reiche et Saulcy, 1857** 3 Detritophages X 

Stenosis sulcata Miller, 1861 3 Detritophages X 

13 


Vegetation structure and environmental gradients 

in the Sallum area, Egypt 

Structure de la végétation et gradients environnementaux 

dans la région de Sallum, Égypte 

Fawzy M. Salama 1 , Monier M. Abd El-Ghani 2-* , Salah M. El Naggar 1 and Khadija A. Baayo 1 

1. Botany Department, Faculty of Science, Assiut University, Assiut 71516, Egypt 

2-*. Corresponding Author; The Herbarium, Faculty of Science, Cairo University, Giza 12613, Egypt; email: elghani@yahoo.com 

Abstract 

We examined vegetation composition and its relation to environmental 

variables in the Sallum area along the mediterranean coastal land 

of Egypt. The study area lies between 25º 09’-25º 35’E and 31º 32’- 

31º 15’ N (about 1700km 2 ), and extends for about 49 km between 

Buqbuq and Sallum on the egyptian-libyan frontier. It is included in 

the semi-desert vegetation zone with an attenuated desert climate. An 

analysis of vegetation along environmental gradients that prevail in 

the study area using the relative importance values of 55 perennials 

in 53 stands, followed by multivariate data analysis was presented. 

Altogether, 111 species (75 perennials and 36 annuals) belonging 

to 92 genera and 34 families of the flowering plants were recorded. 

Asteraceae, Fabaceae, Chenopodiaceae, Poaceae, Brassicaceae, 

Caryophyllaceae, Liliaceae and Zygophyllaceae were the largest 

families, and constitute more than 64 % of the total recorded species. 

Therophytes and chamaephytes were the most frequent, denoting a 

typical desert life-form spectrum. Phytochorological analysis revealed 

that 45.2 % of the studied species were uniregional, of which 23 % 

being native to the saharo-arabian chorotype. It also showed the 

decrease in the numbers of the Mediterranean species and increase 

of the saharo-arabian species along N-S direction form the see-shore 

inwards till the Diffa Plateau. Classification of the vegetation was 

analysed using TWINSPAN technique resulted in the recognition 

of five vegetation groups, each of definite floristic composition and 

environmental characteristics, and could be linked to a specific 

habitat. Haloxylon salicornicum occupied the foot of the Diffa 

Plateau, Haloxylon salicornicum-Thymelaea hirsuta characterised 

the sand plains, Thymelaea hirsuta-Anabasis articulata inhabited 

the non-saline depressions, Haloxylon salicornicum-Atriplex portulacoides 

characterised the saline depressions and Salsola tetrandra-Limoniastrum 

monopetalum on the coastal salt marshes. 

Ordination techniques as DCA and CCA were used to examine the 

relationship between the vegetation and the studied environmental 

parameters: electric conductivity (EC), pH, calcium carbonate, moisture 

content, organic matter, silt, clay, K + , Mg ++ and altitude. CCA 

axis 1 showed significant correlation with clay, moisture content, pH, 

EC, K + , and altitude, while CCA axis 2 was clearly related to pH, 

Mg ++ and altitude. Species richness was correlated with Mg ++ , while 

Shannon index was correlated with Na + and Mg ++ . 

Key-words 

Multivariate analysis, plant distribution, coastal vegetation, diversity, 

plant communities, arid ecosystems 

Résumé 

Nous avons examiné la composition de la végétation et la relation avec 

les différentes variables environnementales dans la région de Sallu, le 

long du littoral méditerranéen de l’Égypte. La zone d’étude se situe 

entre 25º 09’ - 25º 35’ E et 31º 32’ - 31º 15’ N (ca. 1 700 km 2 ), 

et s’étend sur environ 49 km entre Buqbuq et Sallum, à la frontière 

entre l’Égype et la Libye. Cette région s’intègre dans la zone semidésertique 

de végétation, avec un climat désertique atténué. L’analyse 

de végétation a été réalisée en fonction des gradients environnementaux 

locaux, en utilisant les valeurs relatives d’abondance de 55 espèces 

vivaces présentes dans 53 placettes ; ces données ont été traitées par 

analyses statistiques multivariées. Au total, 111 espèces – 75 vivaces 

et 36 annuelles – appartenant à 92 genres et 34 familles ont été 

inventoriées. Les Asteraceae, Fabaceae, Chenopodiaceae, Poaceae, 

Brassicaceae, Caryophyllaceae, Liliaceae et Zygophyllaceae sont les 

familles les plus représentées ; elles regroupent plus de 64 % de toutes 

les espèces enregistrées. Les thérophytes et les chamaephytes sont les 

types biologiques les plus fréquents, ce qui représente un cas classique 

pour la végétation désertique. L’analyse phytochorologique indique 

que 45,2 % des espèces étudiées ont une répartition uni-régionale, 

avec 23 % de végétaux indigènes au chorotype saharo-arabe. Il 

existe une diminution du nombre d’espèces méditerranéennes et une 

augmentation des espèces saharo-arabes selon un gradient nord-sud, 

depuis le bord de mer jusqu’à l’intérieur des terres (plateau de Diffa). 

Le classement de la végétation basé sur la méthode TWINSPAN a 

permis d’identifier 5 groupes de végétation, chacun déterminé par 

une composition floristique et des caractéristiques environnementales 

bien déterminées, liées à un habitat particulier. 

Mots-clés 

Analyses multivariées, distribution des végétaux, végétation côtière, 

diversité, communautés végétales, écosystèmes arides 

15 


◆ FAWZY M. SALAMA, MONIER M. ABD EL-GHANI, SALAH M. EL-NAGGAR, KHADIJA A. BAAYO 

16 

INTRODUCTION 

The Mediterranean coastal land of Egypt extends for 

about 970 km between Sallum on the egyptian-libyan 

border eastward to Rafah on the egyptian-palestinian border, 

with an average width ranging between 15-20 km in 

a N-S direction. It lies within the Mediterranean/Sahara 

regional transition zone, where the vegetation comprises 

floristic elements for both of the mediterranean and 

saharo-arabian regions (White, 1993). Floristically, it 

remains one of the less known territories of the country. 

El Hadidi (2000) distinguished between a Mareotis sector 

which extends between Sallum eastwards to Alexandria, 

where cyrenaican elements are prominent and a sinaitic 

sector extending from Port Said eastwards to Rafah, 

where East Mediterranean taxa prevail. Ecologically, it 

represents the narrow less arid belt of Egypt that can be 

divided into three sections: western, middle and eastern 

(Zahran et al., 1990). The western section extends from 

Sallum eastwards to Abu Qir, near Alexandria, for about 

550 km (fig. 1). 

Physiographically, the western section of the mediterranean 

coastal land can be distinguished into two main 

provinces: an eastern province between Alexandria and 

Ras El-Hikma, and a western province between Ras 

El-Hikma and Sallum (Selim, 1969). One of the salient 

features of the latter province is the dissection of its landscape 

into an extensive system of shallow wadies (gullies, 

sensu El Hadidi, 2000). They draining from the southern 

limestone plateau which lies parallel to the west mediterranean 

coast and reaches a maximum elevation of about 

200 m above sea level at Sallum. The phytosociology and 

vegetation analyses of these wadies were the subject of El 

Hadidi & Ayyad (1975), El Hadidi et al. (1986), El-Kady 

& Sadek (1992), Kamal & El-Kady (1993) and El Garf 

(2003). 

Several earlier extensive studies (1950-1996) depicted 

the vegetation, flora, phytogeography and biodiversity 

of the western section of the mediterranean coastal 

land, amongst others; Tadros & Atta (1958), Tadros & 

El-Sharkawi (1960), Ayyad (1973), Ayyad & El-Kady 

(1982), Shaltout (1985), El-Kady et al. (1995), Ayyad 

Fig. 1. Location map of the study area 

showing the distribution of the studied 

stands along the six transects (T 1 -T 6 ). 


VEGETATION STRUCTURE AND ENVIRONMENTAL GRADIENTS IN THE SALLUM AREA, EGYPT ◆ 

& Fakhry (1996). At least 10 habitat types were recognised 

in this section, which belong to five vegetation 

types. These include the desert vegetation, the littoral salt 

marshes, the coastal sand dunes, the farmland vegetation, 

and the aquatic vegetation (see Zahran & Willis, 1992 

and El Hadidi, 2000 for detailed information). Yet, not 

all parts of that section have been studied in the same 

intense way. Less attention has been paid to the distant 

western part of the mediterranean coast from Sidi Barrani 

to Sallum on the egyptian-libyan frontier (fig. 1), that will 

be here referred to as the Sallum area. It represents the 

distant part of the western mediterranean coastal land of 

Egypt, where the human activities through cultivation, 

grazing and urbanisation were much less pronounced 

than in the other parts of the region. Therefore, the environmental 

correlates of species distribution in that part 

will be highlighted in this study. 

Key questions in the present study were: (1) What are 

the major environmental gradients associated with the 

observed patterns of plant communities? (2) Is the distribution 

of plant communities adequately explained by 

the environmental factors considered or are other factors 

more important in determining species distribution? and 

(3) Which of the studied environmental factors affecting 

the diversity of plant communities inhabiting the study 

area? These objectives were addressed by applying classification 

and ordination techniques to data of species 

composition and environment from the study area. 

STUDY AREA 

The area selected for the present study lies between 

25º 09’-25º 35’ E and 31º 32’-31º 15’ N, extends for about 

49 km from Buqbuq to Sallum on the egyptian-libyan 

frontier with a total area of about 1 700 km 2 (fig. 1). It is 

included in the semi-desert vegetation zone that proposed 

by Bornkamm & Kehl (1990). The coastal plain is very 

narrow or, sometimes, lacking at Sallum. As reported by 

Selim (1969), the entire northern region of the egyptian 

western desert is covered by sedimentary formations 

that range in age from lower Miocene to Holocene. The 

holocene formations comprise lagoon and loamy deposits, 

sand dune accumulations, wadi fillings and limestone 

crusts. Older formations (Pleistocene-Miocene) are of 

limestone or dolomite (dolostone). 

Wickens (1992) proposed that Egypt falls within 

the subtropical desert and mediterranean zonobiomes. 

According to the classification of climatic zones of Egypt 

that suggested Ayyad & Ghabbour (1986) and recently 

modified by El Hadidi (2000), the study area lies within 

the northern arid province that runs parallel to the semiarid 

belt. Comparing climatic characteristics of four 

meteorological stations along the western sector of the 

mediterranean coast (fig. 1): Sallum station, the nearest 

to the study area, with other three are shown in table 1. 

Notably, the wide variation in annual precipitation along 

the E-W direction along the coast (from Alexandria in the 

east to Sallum in the west) was remarkable. 

METHODS 

Vegetation analysis 

During 2002-2003, a quantitative survey of the 

vegetation of the study area was carried out. Six transects 

(T 1 -T 6 ) from the coast till the fringes of the Diffa Plateau, 

were used to sample the vegetation and environmental 

attributes of the study area (fig. 1). Because of the variable 

17 

Station 

Min. 

Temperature (°C) 

Max 

Relative humidity (%) 

Rainfall (mm) 

Sallum 14.0 24.3 65.5 89.8 

Sidi Barrani 15.5 23.3 69.7 149.8 

Mersa Matruh 14.3 24.3 66.0 144.0 

Alexandria 14.9 24.9 65.0 192.1 

Table 1. Long-term annual averages (courtesy of the Egyptian Meteorological Authority) of some climatic factors of Sallum station 

in the study area, and other three stations along the mediterranean coast of Egypt. 



18 

width between the coastal plain and the Diffa Plateau, 

the transects ranged from approximately 2 to 20 km in 

length. These transects included plant communities of the 

major habitats prevailing in the physiographic provinces 

of the region. Stratified random sampling method was 

employed (Greig-Smith, 1983; Ludwig & Reynold, 

1988) within each transect. A total of 53 stands (30 x 

20 m each) were sampled, positioned using GPS model 

Trimble SCOUT M , and distributed along the studied 

transects. In each stand, the plant cover (m/100m) was 

determined using the line-intercept method (Canfield, 

1941). For this purpose, five parallel lines (20 m each) 

were laid across the stand and the intercept lengths (cm) 

of each perennial were summed. Density (individuals/ 

100 m 2 ) and frequency (occurrences/100 sample plots) 

were estimated using a number of randomly located 

sample plots without overlap. The relative importance 

value (IV) for each species in each stand was obtained 

by the sum of its relative density, frequency and cover 

with a maximum value out of 300. Voucher specimens 

of each species were collected and identified by us in 

the Herbaria of Cairo (CAI) and Assiut Universities 

where they are deposited. Taxonomic nomenclature was 

according to Täckholm (1974), Cope & Hosni (1991), 

and Boulos (1995, 1999, 2000 & 2002). 

Soil analysis 

For each sampled stand, three soil samples were collected 

from profiles of 0-15, 15-25 and 25-50 cm depth. 

These samples (300 g each) are then pooled together to 

form one composite sample, air-dried, thoroughly mixed, 

and passes through a 2 mm-sieve to get rid of gravel and 

boulders. The portion finer than 2 mm was kept for physical 

and chemical analysis according to Jackson (1967) 

and Allen & Stainer (1974). Soil texture was determined 

by the Bouyoucos hydrometer method, and the results 

used to calculate the percentages of sand, silt and clay. 

Electric conductivity and pH were evaluated in 1:5 soilwater 

extract using electric conductivity meter and a glass 

electrode pH-meter, respectively. Total carbonate content 

was determined using the volumetric calcimeter (Allison 

& Moodie, 1965), and the Walkely-Black wet combustion 

method (Tan, 1996) using K 2 Cr 2 O 7 and H 2 SO 4 to 

estimate the decomposed fraction of soil organic matter. 

Determination of Na + and K + was carried out by a 

Drlang M 7D flame photometer. Mg ++ was estimated 

using a buck scientific model 210 VGP atomic absorption 

spectrophotometer. 

Data analysis 

In order to obtain an effective analysis of the vegetation 

and related environmental factors, both classification 

and ordination techniques were employed. Two- 

Way INdicator SPecies ANalysis (TWINSPAN) using 

the default settings of the computer program PC-ORD 

for windows version 4.14 (McCune & Mefford, 1999) 

classified a data matrix of 53 stands and 55 perennial 

species using their relative importance values (IV). Only 

species present in at least two stands were included in the 

analysis. TWINSPAN is a divisive hierarchical classification 

method that doubles the number of groups at each 

division (positive and negative groups are formed at each 

dichotomy). The procedures simultaneously classify both 

samples and species directly, constructing an ordered twoway 

table to exhibit the relationship between them clearly 

as possible (Hill, 1979; Økland, 1990). 

The computer program CANOCO 3.12 (ter Braak, 

1987-1992) was used for all ordinations. Rare species 

were downweighted to reduce distortion of the analysis. 

Preliminary analyses were made using Detrended 

Correspondence Analysis (DCA; Hill & Gauch, 1980) 

to check the magnitude of change in species composition 

along the first ordination axis (i.e., gradient length in standard 

deviation (SD)- units). In the present study, DCA 

estimated the gradient length to be larger than 6 SD-units 

for all subset analyses, thus Canonical Correspondence 

Analysis (CCA) was the appropriate ordination method 

to perform direct gradient analysis (ter Braak & Prentice, 

1988). CCA was used to determine the relationships 

between vegetation data and environmental variables 

(Jean & Bouchard, 1993). Ter Braak (1986) suggests 

using DCA and CCA together to see how much of the 

variation in species data is accounted for by the environmental 

data. Due to high inflation factor of total soluble 

salts (TSS), sand, and Na + , they were excluded from 

the CCA analysis. Thus, 10 soil parameters were included: 

electric conductivity (EC), pH, calcium carbonate 

(CaCO 3 ), moisture content (MC), organic matter (OM), 

silt, clay, K + , Mg ++ and altitude (Alt) as a mesologic parameter. 

All the default settings were used for CCA, and 

a Monte Carlo permutation test (99 permutations) was 

used to test for significance of the eigenvalues of the first 

canonical axis. Intra-set correlations from the CCA’s 

were used to assess the importance of the environmental 

variables. The variables in the CCA biplots were represented 

by arrows pointing in the direction of maximum 

variation, with their length proportional to the rate of 

change (ter Braak, 1986). Each arrow determines an axis 



on which the species points can be projected. When plot 

points were projected perpendicularly to the (prolonged) 

arrows, their order represents approximately the ranking 

of weighed averages with respect to the values of the 

factors involved. 

One-way analysis of variance (ANOVA) was applied 

to assess the significance of variations in soil variables 

and diversity indices in relation to the vegetation groups 

identified after TWINSPAN application. Pearson’s product-moment 

correlation coefficient was also used to 

estimate the intercorrelation between the studied soil 

factors. All the statistical analyses were carried out using 

SPSS version 10.0 for Windows. 

Species diversity 

Species diversity within each separated TWINSPAN 

vegetation group was assessed using two different indices 

expressing species richness and diversity. Species richness 

(alpha-diversity) is calculated as the average number of 

species per stand and the species diversity was estimated 

as the Shannon-Wiener index: H’ = Σ i=1 p i log 2 p i , where 

S is the total number of species and p i is the relative 

importance value (IV) of ith species (Whittaker, 1972; 

Pielou, 1975). 

RESULTS 

Taxonomic patterns 

A total of 111 species (36 annuals and 75 perennials) 

belonging to 92 genera and 34 families of the 

vascular plants were recorded (appendix 2). The largest 

families were Asteraceae and Fabaceae (16 for each), 

Chenopodiaceae (10), Poaceae (9), Brassicaceae (7), 

Caryophyllaceae (5), Liliaceae and Zygophyllaceae (4 for 

each). The largest genera include Astragalus (7), Lotus and 

Erodium (3 for each), Launaea, Atriplex, Silene, Medicago, 

Limonium, Asparagus and Asphodelus (2 for each). The 

most common perennials recorded were Haloxylon salicornicum, 

Thymelaea hirsuta, Asphodelus ramosus, Anabasis 

articulata, Atriplex portulacoides, Limoniastrum monopetalum 

and Salsola tetrandra. Each of these species attains a 

maximum importance value (IV) of more than 140 (out 

of 300 for all species in a stand), and a mean of more 

than 60 (table 2). Common but less important perennials 

were Retama raetam, Deverra tortuosa, Lycium europaeum, 

Arthrocnemum macrostachyum, Halocnemum strobilaceum, 

Periploca angustifolia and Zygophyllum album. Common 

annuals include Trigonella stellata, Senecio glaucus, Cotula 

cinerea, Eremobium aegyptiacum, Arnebia decumbens, 

Calendula arvensis, Aizoon hispanicum, Schismus barbatus, 

Erodium laciniatum and Bassia muricata. 

The life-form spectrum of the Sallum area showed 

that the proportion of therophytes (32.4 %) is higher 

than that of other life forms, while the proportions of 

chamaephytes (25.2 %), hemicryptophytes (22.5 %) and 

cryptophytes (13.5 %) were noteworthy (Salama et al., 

2003). The majority of the recorded species belong to the 

mediterranean and saharo-arabian chorotypes. 

Classification of vegetation 

The application of TWINSPAN on the relative 

importance values (IV) of the 55 perennial species 

recorded in 53 sampled stands helped to distinguish 

five vegetation groups (table 2, fig. 2). These groups 

were named after their leading dominant species (those 

have the highest relative IV) as follows: (A) Haloxylon 

salicornicum, (B) Haloxylon salicornicum-Thymelaea 

hirsuta, (C) Thymelaea hirsuta-Anabasis articulata, (D) 

Haloxylon salicornicum-Atriplex portulacoides, and (E) 

Salsola tetrandra-Limoniastrum monopetalum. Each of these 

groups could easily be linked to a habitat type: foot of the 

Diffa Plateau, sand plains, non-saline depressions, saline 

depressions and the coastal salt marshes, respectively. 

Table 3 summarises the mean values and the standard 

deviations of the measured soil variables, and the diversity 

indices in the five groups derived from TWINSPAN. 

The first TWINSPAN dichotomy differentiated the 53 

stands into two main groups according to pH, EC, Na + , 

K + and Mg ++ (p=0.0001). Group E (12 stands) dominated 

by Salsola tetrandra-Limoniastrum monopetalum, 

inhabited the coastal saline depressions was separated 

on the right side of the dendrogram (fig. 2), while the 

left side is still heterogeneous. At the second hierarchical 

level, the inland dry group of stands (41) was split into 

two subgroups related to pH, EC, organic matter, Na + 

and altitude (p=0.0001). Here, another distinct group (A; 

9 stands) dominated by Haloxylon salicornicum found on 

the gravel plains at the foot of Diffa Plateau was also separated. 

Description of each group will be given below. 

Group A. Haloxylon salicornicum group 

The 9 stands of this group were sampled from the 

foot of the Diffa Plateau (fig. 1). On the highly elevated 

and gravely calcareous soil with moderate moisture 

content and the least amounts of organic matter and 

19 



Fig. 2. TWINSPAN dendrogram 

of the 53 studied stands 

of the study area based on their 

importance values. 

A-E are the five separated 

vegetation groups. 

For abbreviation of indicator 

species, see appendix. 

20 

salinity (table 3), sand sheets of Haloxylon salicornicum 

were found. It was differentiated by the growth of shrubs 

of Retama raetam, Lycium europaeum and Farsetia aegyptia, 

and occupied parts of the drainage channels of the 

southern stretches of T 3, T 4 and T 5 where surface deposits 

were deeper. This group had the largest share (23) of 

annuals. Shortly after rainfall, the soil surface supporting 

the sites of this group was covered with a dense vegetation 

of annual species, especially Schismus barbatus, Anthemis 

microsperma, Reichardia tingitana, Brassica tournefortii, 

Medicago laciniata, Cutandia memphitica, Erodium pulverulentum, 

Malva parviflora and Astragalus hamosus. 

Group B. Haloxylon salicornicum- 

Thymelaea hirsuta group 

The landscape of this group was characterized by a 

combination of Haloxylon salicornicum and Thymelaea 

hirsuta found on the sand plains with deep loose soil 

and the lowest levels of moisture content. It represents 

a transitional zone between the non-saline and saline 

depression vegetation groups. This group was differentiated 

by a number of woody species such as Periploca 

angustifolia, Deverra tortuosa, Globularia arabica and Zilla 

spinosa. The herb layer was relatively sparse, and characterized 

by Hordeum leporinum, Asphodelus tenuifolius, 

Bupleurum lancifolium and Astragalus pereginus. The most 

common xerophytic species in the egyptian desert, Sinai, 

and the neighboring arid environments were included in 

this group (Zohary, 1973; Batanouny, 1979; Salama & 

Fayed, 1990; Abd El-Ghani & Amer, 2003). 

Group C. Thymelaea hirsuta- 

Anabasis articulata group 

This vegetation group dominated the non-saline 

depressions with soils of the highest pH values and the 

lowest levels of carbonate content (table 3). Other physical 

soil properties were comparable to those of group B. 

While Gymnocarpos decandrum, Asphodelus ramosus and 

Astragalus siberii dominated the shrub layer, the herb 

layer showed the lowest share of annuals (viz., Centaurea 

glomerata, Louts angustissimus, Asphodelus tenuifolius and 

Hordeum leporinum). 

Group D. Haloxylon salicornicum- 

Atriplex portulacoides group 

This was the largest group of stands (13), and the 

most diversified among the other vegetation groups 

(table 2). It inhabited the saline depressions on fertile 

soils that are rich in their fine sediment contents (silt and 

clay). Relatively high soil salinity favoured the growth of 

some salt-tolerant species as Atriplex portulacoides, Salsola 

tetrandra and Nitraria retusa. The shrub layer was cha- 



TWINSPAN GROUP 

A B C D E 

Group size 9 8 11 13 12 

Total number of species 14 17 16 27 19 

Total number of annuals 23 7 4 9 9 

Haloxylon salicornicum 155 78 36 94 6 

Retama raetam 49 - - - - 

Astragalus sieberi 4 - 8 1 1 

Carthamus glaucus 5 2 - 10 

Hyoscyamus muticus 8 2 1 - - 

Lycium europaeum 17 - 1 4 

Farsetia aegyptiaca 9 - 1 - - 

Periploca angustifolia 7 16 - - - 

Citrullus colocynthis 7 - - - - 

Euphorbia retusa 10 - - - - 

Marrubium alysson 6 - - - - 

Thymelaea hirsuta 5 75 100 25 7 

Deverra tortuosa - 50 2 26 7 

Lygeum spartaum - 1 - - 15 

Globularia arabica - 10 - - - 

Zilla spinosa subsp. biparmata - 15 - - - 

Anabasis articulata 5 15 82 26 - 

Asphodelus ramosus - 24 46 - - 

Carduncellus mareoticus - - 1 3 - 

Echinops spinosus - - 2 1 - 

Gymnocarpos decandrus - - 7 - - 

Atriplex portulacoides - - 3 64 31 

Helianthemum lippii - - - 4 - 

Nitraria retusa - - - 3 - 

Noaea mucronata - - - 2 - 

Verbascum letourneuxii - - - 2 - 

Salsola tetrandra - 1 3 26 66 

Limoniastrum monopetalum - - - - 65 

Halocnemum strobilaceum - - - 1 69 

Suaeda maritima - 1 1 1 13 

Arthrocnemum macrostachyum - - - 1 15 

Sporobolus spicatus - - - 1 9 

Zygophyllum album - - - 2 14 

Frankenia hirsuta - - - - 8 

Limonium pruinosum - - - - 8 

Table 2. Species composition 

of the 53 stands in the six transects, 

arranged in order of occurrence 

in the five TWINSPAN groups (A-E). 

The mean importance value 

(out of 300) rounded to the nearest 

integer is given in each group. 

Entries in bold are indicator 

species in each group. 

See text for explanation. 

21 



22 

Range Mean ± SD 

TWINSPAN vegetation groups 

A B C D E 

F-ratio 

Sand 99.9 – 99.9 99.9 ± 0.08 99.7 ± 0.02 99.9 ± 0.04 99.8 ± 0.07 99.8 ± 0.07 99.9 ± 0.1 2.1 

Silt 0.01 – 0.4 0.07 ± 0.07 0.01 ± 0.02 0.07 ± 0.03 0.08 ± 0.06 0.1 ± 0.06 0.06 ± 0.09 2.5 

Clay 0.01 - 0.1 0.02 ± 0.02 0.01 ± 0.03 0.02 ± 0.01 0.01 ± 0.01 0.02 ± 0.01 0.04 ± 0.03 3.8* 

MC (%) 0.5 – 28.4 3.4 ± 4.6 3.16 ± 3.4 2.12 ± 0.09 2.2 ± 1.70 2.5 ± 1.6 6.7 ± 8.3 2.2 

OM 0.01 - 0.6 0.2 ± 0.01 0.08 ± 0.09 0.19 ± 0.13 0.2 ± 0.14 0.3 ± 0.14 0.2 ± 0.2 3.9* 

Total carbonates 1.8 – 12.5 5.8 ± 2.4 6.9 ± 2.2 5.7 ± 2.30 4.5 ± 1.90 6.0 ± 2.1 5.90 ± 3.1 1.3 

pH 8.2 – 10.0 9.1 ± 0.03 9.2 ± 0.4 9.2 ± 0.15 9.4 ± 0.20 9.3 ± 0.2 8.70 ± 0.1 7.0* 

EC (mS cm -1 ) 0.3 – 4.4 1.1 ± 1.80 0.36 ± 0.03 0.45 ± 0.10 0.41 ± 0.06 0.6 ± 0.2 3.30 ± 3.0 8.9* 

Alt (m) -40 – 100 25.6 ± 27.3 62.0 ± 28.2 31.0 ± 21.8 21 ± 15.3 23.1 ± 21.4 2.0 ± 12.7 12.1* 

Na + 0.02 – 3.8 0.4 ± 0.08 0.04 ± 0.01 0.11 ± 0.08 0.08 ± 0.05 0.2 ±0.15 1.5 ± 1.1 15.7* 

K + (m eq/l) 0.02 - 0.03 0.06 ± 0.05 0.03 ± .007 0.04 ± 0.02 0.04 ± 0.01 0.06 ± 0.02 0.12 ± 0.07 8.4* 

Mg ++ 0.003 -0.03 0.01 ± 0.006 0.005 ± 0.009 0.008 ± 0.003 0.008 ± 0.003 0.009 ± 0.004 0.02 ± 0.008 7.3* 

Species diversity 0.4 – 3.3 1.83 ± 0.07 1.5 ± 0.06 1.70 ± 0.07 1.6 ± 0.6 2.2 ± 0.5 1.9 ± 0.3 2.4 

Species richness 2.0 – 13.0 6.0 ± 2.6 1.9 ± 1.6 6.0 ± 2.4 5.2 ± 1.8 7.1 ± 2.9 6.5 ± 3.4 1.4 

Table 3. Mean values and standard deviations (±) of the environmental variables and altitude in the stands representing the vegetation groups obtained 

by TWINSPAN. MC= Moisture content, OM= Organic matter, EC= Electric conductivity and Alt= Altitude. *= Differences significant at p< 0.05. 

racterised by the growth of Carthamus glaucus, Anabasis 

articulata, Carduncellus mareoticus and Deverra tortuosa. 

Other species showed certain degree of fidelity since 

they did not penetrate to other vegetation groups; such 

as Helianthemum lippii, Noaea mucronata and Verbascum 

letourneuxii. Few annual species (Bassia muricata, 

Astragalus hamosus, Centaurea glomerata and Asphodelus 

tenuifolius) were recorded. 

Group E. Salsola tetrandra- 

Limoniastrum monopetalum 

On muddy fertile saline soil with high Na + , K + and 

Mg ++ , the coastal salt marshes with vegetation characterized 

by the complex Halocnemum strobilaceum, Salsola 

tetrandra and Limoniastrum monopetalum were found, 

indicating the saline nature of this group (12 stands). 

Notably, several halophytic species were recorded such 

as Suaeda maritima, Arthrocnemum macrostachyum, 

Zygophyllum album, Frankenia hirsuta and Limonium 

pruinosum. The herb layer was modestly represented, and 

included amongst others Brassica tournefortii, Centaurea 

glomerata, Astragalus hispidulus and Hordeum leporinum. 

Environmental characteristics of the five vegetation 

groups were summarised in table 3. Of the measured 

parameters, clay, organic matter, pH, electric conductivity, 

altitude, Na + , K + and Mg ++ showed highly significant 

differences between groups. It can be noted that clay, 

moisture content, electric conductivity, K + and Mg ++ 

displayed relatively high values on coastal salt marshes 

(group E), organic matter on the saline depressions 

(group D), and total carbonates and altitude on the foot 

of Diffa Plateau (group A). A remarkable decrease in salinity 

gradient from the coastal salt marshes (group E) to 

the foot of Diffa Plateau (group A) was also noticeable. 

Ordination of stands 

Figure 3 showed the ordination results of the DCA 

analysis of the floristic data set. The 53 stands were 

plotted along axes 1 and 2, and tend to cluster into five 

groups that resulted from TWINSPAN analysis described 

above. The stands were spread out 6.4 S.D units along 

the first axis (eigenvalue= 0.81), expressing the high floristic 

variations among vegetation groups, and indicating 

a complete turnover in species composition took place 

(Hill, 1973). This diagram displayed graphically that 

group B was transitional in its composition between the 

other groups. Stands of group E were separated towards 

and the positive end of DCA axis 1, while those of groups 

A and C were separated out along the other end. DCA 

axis 2 with an eigenvalue of 0.53 and a gradient length of 

3.86 S.D is less important. The species-environment correlation 

(table 4) was also high: 0.83 and 0.55 for DCA 

axis 1 and 2 showing that the species data were related 

to the measured environmental variables. From table 4, 



Fig. 3. DCA ordination diagram 

for the 53 stands on the first 

two axes, with the TWINSPAN 

groups superimposed. 

23 

significant correlations of soil variables with the first three 

DCA axes revealed greater correlations along axis 1 than 

the higher order axis. DCA axis 1 showed significant correlations 

with altitude, EC, K + , moisture content and clay. 

We interpreted DCA axis 1 as an altitude-soil salinity 

gradient. As axis 2 was significantly correlated with total 

carbonates and altitude, DCA axis 2 was interpreted as 

an altitude-carbonate gradient. 

Environment-vegetation-diversity 

relationships 

The successive decrease of the eigenvalues of the first 

three CCA axes (table 4), suggesting a well-structured 

data set. These eigenvalues were somewhat lower than 

for the DCA axes, indicating that important explanatory 

stand variables were not measured and included in the 

analysis or some of the variations was not explained 

by environmental variables (Franklin & Merlin, 1992; 

McDonald et al., 1996). However, the species-environment 

correlations were higher for the first three canonical 

axes, explaining 67.3 % of the cumulative variance. These 

results suggested a strong association between vegetation 

and the measured soil parameters presented in the biplot 

(Jongman et al., 1987). From the intra-set correlation of 

the soil factors with the first three axes of CCA shown in 

table 4, it can be noted that CCA axis 1 was correlated 

to clay, moisture content, pH, EC, K + , and altitude. This 

fact becomes more clearly in the biplot (fig. 4). A test for 

significance with an unrestricted Monte Carlo permutation 

test (99 permutations) found the F-ratio for the 

eigenvalue of axis 1 and the trace statistics to be significant 

(p


DCA AXIS 

CCA AXIS 

1 2 3 1 2 3 

Eigenvalues 0.81 0.55 0.32 0.65 0.38 0.33 

Species-environment correlations 0.83 0.55 0.70 0.93 0.85 0.78 

Silt -0.24 -0.15 0.09 0.3 -0.001 0.02 

Clay -0.59 * -0.03 0.01 0.74 * 0.04 0.03 

MC -0.60 * 0.01 -0.21 0.71 * -0.1 -0.27 

OM -0.32 -0.1 0.04 0.37 -0.08 0.18 

Total Carbonates -0.11 0.39 * -0.21 -0.2 0.10 -0.32 

pH 0.51 -0.22 0.32 -0.57 * -0.22 * 0.41 * 

EC -0.71 * 0.04 -0.05 0.86 * -0.15 -0.09 

K + -0.67 * 0.12 -0.1 0.80 * 0.12 -0.13 

Mg ++ -0.51 * 0.02 -0.42 0.41 0.65 * -0.18 

Altitude 0.56 * 0.32 * -0.05 -0.50 * -0.38 * -0.34 * 

24 

Table 4. Comparison of the results of ordination for the first three axes of DCA and CCA. Intra-set correlations of the environmental variables and altitude, 

together with eigenvalues and species-environment correlation coefficients. *= Significant at p


the relative positions of species and sites along the most 

important ecological gradients. Both ordination techniques 

clearly indicated that fine soil sediments, moisture 

content, pH, electric conductivity, altitude, relative concentrations 

of K + and Mg ++ were the most important 

parameters for the distribution of the vegetation pattern 

in the present study. The role of these factors in 

delimiting plant communities in the western desert has 

been stressed by many authors (Ayyad, 1976; Kamal & 

El-Kady, 1993). The percentage of surface sediments of 

different size classes determines the spatial distribution of 

soil moisture (Yair & Danin, 1980), as shown for other 

desert ecosystems in Egypt by Sharaf El Din & Shaltout 

(1985), Abd El-Ghani (1998, 2000a) and in Saudi Arabia 

by El-Demerdash et al. (1994). 

The distribution of species in saline and marshy 

habitat relates to salinity in many arid regions has been 

discussed by several authors (Kassas, 1957; Ungar, 1968 

and Maryam et al., 1995). Ungar (1974) indicated that 

the distribution of halophytes in the United States is 

mainly dependent on the salinity gradient, while local 

climate, topography, soil moisture and biotic factors are 

less important. Ragonese & Covas (1947) described the 

interrelation of the salinity gradient and vegetation in the 

northern Argentinian saltmarshes. Abu-Ziada (1980) and 

Abd El-Ghani (2000b) also noted strong relationships 

between the vegetation pattern and the soil moisture-salinity 

gradients in the oases of the western desert of Egypt. 

When studying the saltmarsh communities of the western 

mediterranean coastal desert, Ayyad & El-Ghareeb 

(1982) pointed out that salinity, the concentration of different 

ions and the periodical variation in the water table 

determine the distribution of species and the differences 

between communities. They also conclude that the saltmarsh 

vegetation in this part of the country represents a 

transition from the western communities in North Africa 

and those characteristics of the eastern mediterranean 

region. In their account of the northern and eastern 

mediterranean coastal saltmarshes, Zahran et al. (1996) 

demonstrated the distribution of some halophytic species 

as best correlated along a gradient of a dozen of soil variables, 

the most important were salinity, moisture content, 

soil texture, organic matter and calcium carbonate. In the 

present study, the gradient in soil salinity and its variation 

from one habitat to another was the primary determinant 

of the plant community composition. 

The plant life in the study area was restricted to 

microenvironments (as in wadies, runnels and depressions) 

where runoff water collects and provides sufficient 

moisture for plant growth. This was described by Monod 

(1954) as “végétation contractée”, runoff desert (sensu 

Zohary, 1962), and restricted desert (sensu Walter, 1963). 

25 

Fig. 4. CCA biplot of axes 1 

and 2 showing the distribution 

of the 53 stands with their 

TWINSAPN groups 

and soil variables. 

For abbreviations, see table 2. 



Diversity indices 

26 

Table 5. Spearman rank correlations (r) between the diversity indices 

and the studied environmental variables and altitude in the study area. 

*= Significant at p


References 

ABD EL-GHANI, M.M., 1998. Environmental correlates of species 

distribution in arid desert ecosystems of eastern Egypt. 

Journal of Arid Environments, 38: 297-313. 

ABD EL-GHANI, M.M., 2000a. Floristics and environmental relations 

in two extreme desert zones of western Egypt. Global 

Ecology and Biogeography, 9: 499-516. 

ABD EL-GHANI, M.M., 2000b. Vegetation composition of egyptian 

inland salt marshes. Botanical Bulletin of Academia Sinica, 

41: 305-314. 

ABD EL-GHANI, M.M. & AMER, W.M., 2003. Soil-vegetation relationships 

in a coastal desert plain of southern Sinai. Journal 

of Arid Environments, 55: 607-628. 

ABDEL-RAZIK, M., ABDEL-AZIZ, M. & AYYAD, M., 1984. 

Environmental gradients and species distribution in a 

transect at Omayed (Egypt). Journal of Arid Environments, 

7: 337-352. 

ABU-ZIADA, M.E.A., 1980. Ecological studies on the flora of 

Kharga and Dakhla oases of the western desert of Egypt. 

Ph.D. Thesis, Faculty of Science, Mansoura Univ. 342 pp. 

ALLEN, M.M. & STAINER, S.T., 1974. Chemical analysis of 

Ecological Materials. Blackwell Scientific Publications, 

London. 565 pp. 

ALLISON, L.E. & MOODIE, C.D., 1965. Carbonates. In: Black, C.A. 

et al. (eds), Methods of Soil Analysis, Part 2, p. 1379. American 

Society of Agronomy Incorporation, Madison, USA. 

AYYAD, M. A., 1973. Vegetation and environment of the western 

mediterranean coastal land of Egypt. I. The habitat of sand 

dunes. Journal of Ecology, 61: 509-523. 

AYYAD, M., 1976. The vegetation and environment of the western 

mediterranean coastal land of Egypt. II. The habitat of 

non-saline depressions. Journal of Ecology, 62: 439-456. 

AYYAD, M.A. & EL-GHAREEB, R.E., 1982. Salt marsh vegetation 

of the western mediterranean desert of Egypt. Vegetatio, 49: 

3-19. 

AYYAD, M.A. & EL-KADY, H.F., 1982. Effect of protection and 

controlled grazing on the vegetation of mediterranean desert 

ecosystem in Northern Egypt. Vegetatio, 49: 129-139. 

AYYAD, M.A. & FAKHRY, A.M., 1996. Plant biodiversity in the 

western mediterranean desert of Egypt. Verhandlungen der 

Gesellschaft für Ökologie, 25: 65-76. 

AYYAD, M.A. & GHABBOUR, S.I., 1986. Hot Deserts of Egypt and 

Sudan. In: Evenari, M. et al. (eds.), Ecosystems of the World, 

12B, Hot Deserts and Arid Shrublands. Elsevier, Amsterdam, 

pp. 149-202. 

BATANOUNY, K.H., 1979. The desert vegetation in Egypt. Cairo 

University African Studies Review, Special Publication, 1: 9-37. 

BORNKAMM, R. & KEHL, H., 1985. Pflanzengeographische 

zonen in der Marmarika (Nordwest-Egypten). Flora, 176: 

141-151. 

BORNKAMM, R. & KEHL, H., 1990. The plant communities of the 

western desert of Egypt. Phytocoenologia, 19: 149-231. 

BOULOS, L., 1995. Flora of Egypt. Check list. Al Hadara Publishing, 

Cairo. 283 pp. 

BOULOS, L., 1999. Flora of Egypt. Vol. 1. Azollaceae-Oxalidaceae. 

Al Hadara Publishing, Cairo. 417 PP. 

BOULOS, L., 2000. Flora of Egypt. Vol. 2. Geraniaceae-Bora-ginaceae. 

Al Hadara Publishing, Cairo. 352 pp. 

BOULOS, L., 2002. Flora of Egypt. Vol. 3. Verbenaceae-Compositae. 

Al Hadara Publishing, Cairo. 372 pp. 

BRAUN-BLANQUET, J., 1949. Premier apercu phytosociologique 

du Sahara Tunisien. Mémoire hors-series de la Société d’histoire 

naturelle d’Afrique du Nord, 2 : 1-16. 

CANFIELD, R., 1941. Application of the line interception method 

in sampling range vegetation. Journal of Forestry, 39: 288- 

294. 

COPE, T.A. & HOSNI, H.A., 1991. A Key to Egyptian Grasses. 

Royal Botanic Gardens Kew and Cairo University Herbarium, 

London. 75 pp. 

EL-DEMERDASH, M., HEGAZY, A.K. & ZILAY, A.M., 1994. 

Distribution of the plant communities in Tihamah coastal 

plains of Jazan region, Saudi Arabia. Vegetatio, 112: 141- 

151. 

EL GARF, I.A., 2003. Studies on the shallow wadies of the 

Mareotis sector of the mediterranean coastal land of Egypt. 

I. Floristic features of wady Halazeen. Bulletin of the Faculty 

of Science, Assiut University, 32 (1-D): 61-71. 

EL-GHAREEB, R., 1990. A phytosociological study of Zygophyllum 

album L. in the western mediterranean coast of Egypt. Journal 

of the University of Kuwait (Science), 17: 143-155. 

EL-GHAREEB, R. & HASSAN, I.A., 1989. A phytosociological study 

on the inland desert plateau of the western mediterranean 

desert of Egypt at El-Hammam. Journal of Arid Environments, 

17: 13-21. 

EL HADIDI, M.N., 2000. Geomorphology, climate and phytogeographic 

affinities. In: El Hadidi, M.N. & Hosni, H.A. (eds.), 

Flora Aegyptiaca, vol. 1, pp. 1-25. The Palm Press and The 

Cairo University Herbarium, Cairo, 187 pp. 

EL HADIDI, M.N. & AYYAD, M.A., 1975. Floristic and ecological 

features of Wadi Habis (Egypt). La Flore du bassin méditerranéen 

: essai de systématique synthétique, Colloques internationaux 

du CNRS, 235 : 247- 258. 

EL HADIDI, M.N., ABD EL-GHANI, M.M., SPRINGUEL, I. & 

HOFFMAN, M.A., 1986. Wild barley Hordeum spontaneum L. 

in Egypt. Biological Conservation, 37: 291-300. 

EL-KADY, H.F., AYYAD, M. & BORNKAMM, R., 1995. Vegetation 

and recent land-use history in the desert of Maktala, Egypt. 

Advances in GeoEcology, 28: 109-123. 

EL-KADY, H.F. & SADEK, L., 1992. On the phytosociology of 

wadi Um Ashtan in the western desert of Egypt. Feddes Repertorium, 

103: 269-277. 

27 



28 

FRANKLIN, J. & MERLIN, M., 1992. Species-environment patterns 

of forest vegetation on the uplifted reef limestone of 

Atiu, Managia, Ma’uke and Miti’aro, Cook Islands. Journal 

of Vegetation Science, 3: 3-14. 

GREIG-SMITH, P., 1983. Quantitative Plant Ecology. Third Edition. 

Blackwell Scientific Publications, London, 359 pp. 

HILL, M.O., 1973. Diversity and evenness: A unifying notation 

and its consequences. Ecology, 54: 427-432. 

HILL, M.O., 1979. TWINSPAN-A FORTRAN program for 

arranging multivariate data in an ordered two-way table of 

classification of individuals and attributes. Ithaca, NY, Cornell 

University. 90 pp. 

HILL, M.O. & GAUCH, H.G., 1980. Detrended Correspondence 

Analysis: An improved ordination technique. Vegetatio, 42: 

47-58. 

JACKSON, M.L., 1967. Soil chemical analysis-Advanced course. 

Published by the Washington Department of Soil Sciences. 

498 pp. 

JONGMAN, R.H., TER BRAAK, C.J.F. & VAN TONGEREN, O.F.G., 

1987. Data analysis in community and landscape ecology. 

Pudoc (Wageningen), The Netherlands. 299 pp. 

KAMAL, S.A. & EL-KADY, H.F., 1993. Vegetation analysis of 

some wadis in the egyptian mediterranean desert. Feddes 

Repertorium, 104: 537-545. 

KASSAS, M., 1957. On the ecology of the Red Sea coastal land. 

Journal of Ecology, 45: 187-203. 

KASSAS, M., 1966. Plant life in deserts. In: Hills, E.S. (ed.), 

Arid Lands, pp. 145-180. London, Mathuen/Paris, Unesco, 

461 pp. 

KILLIAN, C. & LEMÉE, G., 1948. Étude sociologique, morphologique 

et écologique de quelques halophytes sahariens. Revue 

de Génétique et de Botanique, 55 : 376-402. 

LUDWIG, J.A. & REYNOLD, J.F., 1988. Statistical Ecology. John 

Wiley, New York, 337 pp. 

MAIRE, R. & WEILLER, M., 1947. Remarque sur la flore et la végétation 

de la Tripolitaine et de la Cyrénaïque septentrionales. 

Recueil de Travaux de l’Institut de botanique, Montpellier, 3 : 

1-17. 

MARYAM, H., ISMAIL, S., ALAA, F. & AHMED, R., 1995. Studies 

on growth and salt regulation in some halophytes as influenced 

by edaphic and climatic conditions. Pakistan Journal of 

Botany, 27: 151-163. 

MCCUNE, B. & MEFFORD, M.J., 1999. PC-ORD for windows. 

Multivariate Analysis of Ecological Data. Version 4.14. User’s 

Guide. MjM Software, Oregon, USA. 237 pp. 

MCDONALD, D.J., COWLING, R.M. & BOUCHER, C., 1996. 

Vegetation-environment relationships on a species-rich 

coastal mountain range in the fynobs biome (South Africa). 

Vegetatio, 123:165-182. 

MIGAHID, A.M., BATANOUNY, K.H. & ZAKI, M.A.F., 1971. 

Phytosociological and ecological study of a sector in the 

mediterranean coastal region in Egypt. Vegetatio, 23: 113- 

134. 

MONOD, TH., 1954. Mode contracté et diffus de la végétation 

saharienne. In: Cloudsley-Thompson, J.L. (ed.), Biology of 

Desert, pp. 35-44. Institute of Biology, London. 224 pp. 

NAVEH, H. & WHITTAKER, R.H., 1979. Structural and floristic 

diversity of shrublands and woodlands in northern Israel and 

other Mediterranean areas. Vegetatio, 41: 171-190. 

NAVIKOFF, G., 1961. Contribution a l’étude des relations entre le 

sol et la végétation halophile de la Tunisie. Annales de l’Institut 

national de la recherche agronomique de Tunisie, 34 : 90-114. 

ØKLAND, R.H., 1990. Vegetation ecology: theory, methods and 

applications with reference to Fennoscandia. Sommerfettia 

Supplement, 1: 1-216. 

PIELOU, E.C., 1975. Ecological Diversity. Wiley, New York. 165 pp. 

PIGNATTI, E. & PIGNATTI, S., 1989. Life forms and phytogeographical 

affinities of the Canarian flora. Flora, 183: 87-95. 

QUÉZEL, P., 1978. Analysis of the flora of mediterranean and 

saharan Africa. Annals of the Missouri Botanical Garden, 65: 

479-534. 

RAGONESE, A. & COVAS, G., 1947. The halophilous flora of 

southern Santa Fe province (Argentina). Darwiniana, 7: 40- 

496. 

SALAMA, F.M., ABD EL-GHANI, M.M., EL-NAGGAR, S.M. & 

BAAYO, K.A., 2003. Floristic composition and chorological 

analysis of the Sallum area, west mediterranean Egypt. 

Journal of Union Arab Biologists Cairo, 13 (B): 27-47. 

SALAMA, F.M. & FAYED, A.A., 1990. Phytosociological study 

on the deltaic part and the principal channel of wadi Qena, 

Egypt. Feddes Repertorium, 101: 89-96. 

SELIM, A.A., 1969. Geology of El-Salum Area, western Mediterranean 

coastal zone, UAR. Ph.D. Thesis, Alexandria 

University. 

SHALTOUT, K.H., 1985. On the diversity of the vegetation in the 

western mediterranean coastal region of Egypt. Proceedings 

of the Egyptian Botanical Society, 4: 1355-1376. 

SHALTOUT, K.H. & EL-GHAREEB, R., 1992. Diversity of the 

salt marsh plant communities in the western mediterranean 

region of Egypt. Journal of the University of Kuwait (Science), 

19: 75-84. 

SHARAF EL DIN, A. & SHALTOUT, K.H., 1985. On the 

phytosociology of Wadi Araba in the eastern desert of Egypt. 

Proceedings of the Egyptian Botanical Society, 4: 1311-1325. 

SIMMONEAU, P., 1954. La végétation de sols salés d’Oranie. Les 

groupements à Atriplex dans les plaines sublittorales. Annales 

d’Agronomie, 2: 226-257. 

TADROS, T.M. & ATTA, B.M., 1958. Further contribution to 

the sociology and ecology of halophilous communities of 

Mareotis (Egypt). Vegetatio, 8: 137-160. 

TADROS, T.M. & EL-SHARKAWI, H.M., 1960. Phytosociological 

and ecological studies on the vegetation of Fuka-Ras El Hekma 



area. II. Consistency and homogeneity of the open desert communities. 

Bulletin de l’Institut du Désert d’Egypte, 10: 61–63. 

TÄCKHOLM, V., 1974. Students’ Flora of Egypt (2 nd Edn.). Cairo 

University (Publ.) & Cooperative Printing Company, Beirut. 

888 pp. 

TAN, K. H., 1996. Soil Sampling, Preparation and Analysis. Marcel 

Dekker Inc. New York, USA. 

TER BRAAK, C.J.F., 1986. Canonical correspondence analysis: a 

new eigenvector technique for multivariate direct gradient 

analysis. Ecology, 67: 1167-1179. 

TER BRAAK, C.J.F., 1987-1992. Canoco-a fortran program for 

Canonical Community Ordination. Micro-computer Power, 

Ithaca, NY. 

TER BRAAK, C.J.F. & PRENTICE, I.C., 1988. A theory of gradient 

analysis. Advances in Ecological Researches, 18: 271-317. 

TIELBÖRGER, K., 1997. The vegetation of linear desert dunes in 

the north-western Negev, Israel. Flora, 192: 261-278. 

UNGAR, I.A., 1968. Species-soil-relationships on the great salt 

plains of northern Oklahoma. American Midland Naturalist, 

80: 392-406. 

UNGAR, I., 1974. Halophyte communities of Park county, 

Colorado. Bulletin of the Torrey Botanical Club, 101: 145-152. 

WALTER, H., 1963. Water supply of desert plants. In: Rutter, 

A.J. & Whitehead, E.H. (eds.), The Water Relations of Plants, 

pp. 199-205. Blackwell, London. 394 pp. 

WHITE, F., 1993. The AETFAT chorological classification of 

Africa: History, methods and applications. Bulletin du Jardin 

botanique national de Belgique, 62: 225-281. 

WHITE, F. & LÉONARD, J., 1991. Phytogeographical links between 

Africa and southwest Asia. Flora et Vegetatio Mundi, 9: 229- 

246. 

WHITTAKER, R.H., 1972. Evolution and measurement of species 

diversity. Taxon, 21: 213-251. 

WICKENS, G.E., 1992. Arid and semiarid ecosystems. Encyclopedia 

of Earth System Science, 1: 113-118. 

YAIR, A. & DANIN, A., 1980. Spatial variation as related to the 

soil moisture regime over an arid limestone hillside, northern 

Negev, Israel. Œcologia, 47: 83-88. 

ZAHRAN, M.A., EL-DEMERDASH, M.A. & MASHALY, I.A., 1990. 

Vegetation types of the deltaic mediterranean coast of Egypt 

and their environment. Journal of Vegetation Science, 1: 301- 

310. 

ZAHRAN, M.A., MURPHY, K.J., MASHALY, I.A. & KHEDR, A.A., 

1996. On the ecology of some halophytes and psammophytes 

in the mediterranean coast of Egypt. Verhandlungen der 

Gesellschaft für Ökologie, 25: 133-146. 

ZAHRAN, M.A. & WILLIS, A.J., 1992. The Vegetation of Egypt. 

Chapman & Hall, London. 424 pp. 

ZOHARY, M., 1962. Plant Life in Palestine. Ronald Press, New 

York. 262 pp. 

ZOHARY, M., 1973. Geobotanical Foundations of the Middle East. 

Vols 1-2. Gustav Fischer Verlag, Stuttgart.739 pp. 

29 

APPENDIX 

Species 

Anabasis articulata (Forssk.) Moq. 

Arthrocnemum macrostachyum (Moric.) K. Koch 

Atriplex portulacoides L. 

Deverra tortuosa (Desf.) DC. 

Halocnemum strobilaceum (Pall. ) M. Bieb. 

Helianthemum lippii (L.) Dum. Cours. 

Lycium europaeum L. 

Retama raetam (Forssk.) Webb & Berthel. 

Salsola tetrandra Forssk. 

Thymelaea hirsuta (L.) Endl. 

Zygophyllum album L. 

Abbreviations 

Aa 

Am 

Ap 

Dt 

Hs 

Hl 

Ll 

Rr 

St 

Th 

Za 

Appendix 1. Names and abbreviations of the indicator 

species displayed in figure 2. First letter for the genus 

and the second for the species. 



30 

Family Species Life form Chorotype 

Amaryllidaceae Pancratium maritimum L. Cr MED 

Apiaceae 

Bupleurum lancifolium Hornem. Th MED+IT 

Deverra tortuosa (Desf.) DC. Ch SA 

Asclepiadaceae Periploca angustifolia Labill. NPh SA+SZ 

Anthemis microsperma Boiss. & Kotschy Th MED 

Artemisia judaica L. Ch SA 

Atractylis carduus (Frossk.) C. Chr. H MED+SA 

Carduncellus mareoticus (Delile) Hanelt Th SA 

Carthamus glaucus M. Bieb. Th MED 

Centaurea glomerata Vahl Th SA 

Chiliadenus candicans (Delile) Brullo H MED+IT 

Asteraceae 

Echinops sipnosus L. H IT 

Hyoseris radiata L. subsp. graeca Halácsy Cr MED 

Hyoseris scabra L. Cr MED 

Koelpinia linearis Pall. Th SA+IT 

Launaea nudicaulis (L.) Hook. F. H SA+SZ+IT 

Launaea fragilis (Asso) Pau H SA 

Phagnalon barbeyanum Asch. & Schweinf. Ch SA 

Reichardia tingitana (L.) Roth Th SA+IT 

Scorzonera undulata Vahl Cr MED+IT 

Arnebia decumbens (Vent) Coss. & Kralik Th SA+IT 

Boraginaceae 

Echiochilon fruticosum Desf. Ch SA 

Heliotropium lasiocarpum Fisch. & C.A. Mey. Ch MED+IT+ES 

Brassica tournefortii Gouan Th MED+SA 

Carrichtera annua (L.) DC. Th SA 

Enarthrocarpus lyratus (Forssk.) DC. Th MED 

Brassicaceae 

Erucaria microcarpa Boiss. Ch MED+IT 

Farsetia aegyptia Turra Ch SA+SZ 

Sisymbrium irio L. Th MED+SA+IT 

Zilla spinosa (L.) Prantl. subsp. biparmata (O.E.Schulz) Maire & Weiller Ch SA 

Gymnocarpos decandrus Forssk. Ch MED+SA 

Herniaria hemistemon J. Gay H SA 

Caryophyllaceae 

Paronychia arabica (L.) DC. Th SA 

Silene succulenta Forssk. H MED 

Silene vivianii Steud. Th SA 

Anabasis articulata (Forssk.) Moq. Ch SA+IT 

Arthrocnemum macrostachyum (Moric.) K. Koch Ch MED+SA 

Atriplex halimus L. NPh MED+IT 

Atriplex portulacoides L. Ch COSM 

Chenopodiaceae 

Bassia muricata (L.) Asch. Th SA+IT 

Halocnemum strobilaceum (Pall.) M. Bieb. Ch MED+SA+IT 

Haloxylon salicorincum (Moq.) Bunge ex Boiss. Ch SZ 

Noaea mucronata (Forssk.) Asch. & Schweinf. Ch MED+IT 

Salsola kali L. Th PAL 

Suaeda pruinosa Lange Ch MED 



Cistaceae Helianthemum lippii (L.) Dum. Cours. Ch SA+SZ 

Convolvulaceae Convolvulus arvensis L. H COSM 

Cucurbitaceae Citrullus colocynthis (L.) Schrad. H SA 

Euphorbiaceae 

Euphorbia dendroides L. Ch MED 

Euphorbia retusa Forssk. H SA 

Argyrolobium uniflorum (Decne.) Jaub. & Spach. Ch MED+SA+SZ 

Astragalus caprinus L. H SA 

Astragalus hamosus L. Th MED+IT 

Astragalus hispidulus DC. Th SA 

Astragalus peregrinus Vahl Th SA 

Astragalus schimperi Boiss. Th SA 

Astragalus sieberi DC. Ch SA 

Astragalus spinosus (Forssk.) Muschl. Ch SA+IT 

Fabaceae 

Lotus angustissimus L. Th MED+ES 

Lotus creticus L. H MED 

Lotus polyphllos E.D. Clarke H MED 

Medicago coronata (L.) Bartal. Th MED+IT 

Medicago laciniata (L.) Mill. Th SA 

Ononis vaginalis Vahl Ch SA+IT 

Retama raetam (Forssk.)Webb & Berthel. NPh MED+SA+IT 

Trigonella stellata Forssk. Th SA+IT 

Frankeniaceae Frankenia hirsuta L. H MED+IT+ES 

Erodium crassifolium L’Hér. Cr SA 

Geraniaceae 

Erodium laciniatum (Cav.) Willd. Th MED 

Erodium pulverulentum (Cav.) Wiild. Th SA 

Globulariaceae Globularia arabica Jaub. & Spach. Ch MED+SA 

Juncaceae 

Juncus acutus L. Cr COSM 

Juncus rigidus Desf. Cr MED+IT+ES 

Ajuga iva (L.) Schreb. H MED 

Lamiaceae 

Marrubium alysson L. Th MED+SA 

Salvia lanigera Poir. H MED+SA 

Asparagus aphyllus L. Cr MED 

Liliaceae 

Asparagus stipularis Forssk. Cr MED+SA 

Asphodelus ramosus L. Cr MED 

Asphodelus tenuifolius Cav. Th MED+SA+SZ 

Iridaceae Gynandriris sisyrinchium (L.) Parl. Cr MED 

Malvaceae Malva parviflora L. Th MED 

Neuradaceae Neurada procumbens L. Th SA 

Plantaginaceae Plantago albicans L. H MED+SA 

Limoniastrum monopetalum (L.) Boiss. Ch MED 

Plumbaginaceae 

Limonium narbonense Mill. H MED 

Limonium pruinosum (L.) Chaz. H SA 

31 



32 

Aeluropus lagopoides (L.) Trin. ex Thwaites Cr MED+IT 

Ammophila arenaria (L.) Link Cr MED 

Cynodon dactylon (L.) Pers. Cr COSM 

Cutandia memphitica (Sperng.) K. Richt. Th MED+SA+IT 

Poaceae 

Hordeum murinum L. subsp. leporinum (Link) Arcang. Th MED+IT 

Lygeum spartum Loefl. ex L. Cr MED 

Schismus barbatus (L.) Thell. Th SA+IT 

Sporobolus spicatus (Vahl) Kunth Cr MED+SA+SZ 

Stipa parviflora Desf. H MED+IT 

Resedaceae Reseda alba L. Th MED+IT 

Rubiaceae 

Galium tricornutum Dandy Th MED+IT+ES 

Crucianella maritima L. H MED 

Rutaceae Haplophyllum tuberculatum (Forssk.) Juss. H SA 

Santalaceae Thesium humile Vahl P MED 

Scrophulariaceae 

Kickxia aegyptiaca (L.) Nábelek Ch SA 

Verbascum letourneuxii Asch. & Schweinf. H MED 

Solanaceae 

Hyoscyamus muticus L. H SA+IT+SZ 

Lycium europaeum L. NPh MED+SA 

Tamaricaceae Reaumuria hirtella Janb. & Spach. Ch MED+IT 

Thymelaeaceae Thymelaea hirsuta (L.) Endl. NPh MED+SA 

Fagonia microphlla Pomel Ch SA+IT 

Zygophyllaceae 

Nitraria retusa (Forssk.) Asch. NPh SA+SZ 

Peganum harmala L. H MED+SA+IT 

Zygophyllum album L.f. Ch MED+SA 

Appendix 2. Distribution of plant species recorded in the Sallum area, together with their life forms and chorotypes. 

Chorotype abbreviations: MED= Mediterranean, SA= Saharo-Arabian, IT = Irano-Turanian, SZ= Sudano-Zambezian, 

ES= Euro-Siberian, COSM= Cosmopolitan and PAL= Palaeotopical. Life-form Abbreviations: Th= Therophytes, Ch= Chamaephytes, 

H= Hemicryptophytes, Cr= Cryptophytes, P= Parasites and NPh= Nano-Phanerophytes. 


Phenological patterns of ground spiders (Araneae, Gnaphosidae) 

on Crete, Greece 

Phénologie des araignées édaphiques (Araneae, Gnaphosidae) 

en Crète, Grèce 

M. Chatzaki 1-2 , G. Markakis 3 & M. Mylonas 1-2 

1. Natural History Museum of Crete, University of Crete, Irakleio, Greece 

2. Department of Biology, University of Crete, Irakleio, Greece 

3. Technological Education Institute of Crete, Irakleio, Greece 

Correspondence: M. Chatzaki, Natural History Museum of Crete, University of Crete, PO Box 2208, 71409 Irakleio, Crete, Greece. 

E-mail: mchatzak@nhmc.uoc.gr 

33 

Summary 

The phenology of one of the dominant families of ground spiders 

of Greece and along the whole Mediterranean, the Gnaphosidae, 

is presented. Phenological patterns of the 11 most common and 

abundant species of this family are compared in relation to each 

other and between 17 sites situated on the island of Crete, Greece. A 

major difference between lowland and high elevation sites is recorded 

for some species. This difference is attributed to a restriction of 

activity and/or a shifting of one to two months towards the summer 

and early autumn. Based on the similarities which exist among 

species a model that describes phenological patterns of ground spiders 

in Mediterranean ecosystems is proposed. Most species present 

single peaks of activity during the dry period, from mid spring to 

mid autumn, which lead to a spectrum of high activity within this 

period. Most Gnaphosidae species on Crete are considered to be 

high competitors, which can very efficiently take advantage of the 

favourable period for this area. Based on their phenology, it is conjectured 

that they either have annual biological cycles or that they 

may produce two generations per year, although, in other latitudes, 

the same species may have biennial cycles. The flexibility of species, 

as far as their presence in both time and space is concerned, reveals 

the combined effect of their own physiological tolerance and the 

environmental heterogeneity in their niche definition. 

Key-words 

Ground spiders, Gnaphosidae, Crete, Greece, Mediterranean, 

phenology 

Résumé 

La phénologie des araignées Gnaphosidae, une des familles 

dominantes en Crète et dans toute la Méditerranée est analysée. 

La phénologie des onze espèces les plus communes est comparée et 

leur stabilité entre les 17 sites de rêcolte est testée. La plupart des 

espèces présentent un seul sommet d’activité pendant la période 

sèche d’avril à octobre, formant un spectre de grande activité tout 

au long de cette période. Une différence entre les régions de plaine 

et de montagne est notée pour quelques espèces. Cette différence est 

attribuée à une diminution de l’activité et/ou un décalage d’un ou 

deux mois vers l’été et le début de l’automne. Basé sur les similarités 

qui existent entre les espèces, un modèle définissant les combinaisons 

phénologiques des araignées du sol dans les écosystèmes méditerranéens 

est proposé. La plupart des Gnaphosidae de Crète sont considérées 

comme de grands compétiteurs qui peuvent tirer avantage de 

cette période favorable. Elles présentent soit des cycles annuels, soit 

deux générations par an, alors que sous d’autres latitudes, les mêmes 

espèces ont un cycle biannuel. La fléxibilité des espèces révèle l’effet 

combiné de leur propre tolérance physiologique et de l’hétérogénéité 

environnementale de leur niche écologique. 

Mots-clés 

Araignées du sol, Gnaphosidae, Crète, Grèce, Méditerranée, 

phénologie 


◆ M. CHATZAKI, G. MARKAKIS & M. MYLONAS 

34 

INTRODUCTION 

Seasonality in climate is reflected in the activity of 

organisms. Varying activity of animals between seasons 

is interrelated (in the sense of mutual causality) with 

the coexistence of groups sharing the same or similar 

food resources in the same environment (Tretzel, 1954; 

Kuenzler, 1958; Williams, 1962; Breymeyer, 1966; 

Dondale et al., 1972; Milner, 1988; Pianka, 1994; Majadas 

& Urones, 2002). In Mediterranean climates seasonality 

is very clear. Organisms are confronted with two periods 

of stress (Mitrakos, 1980): one in the winter, due to 

low temperatures, and one in the summer, due to high 

drought. The two other seasons – autumn and spring – 

are the transitional phases between the above seasons. 

Spiders are very tolerant to both kinds of climatic 

stress, i.e. high/low temperatures and aridity, and that is 

why they constitute the main group of ground predators 

throughout the year in most types of Mediterranean 

habitats (Paraschi, 1988; Radea, 1993). Although 

maximum species richness and activity in all studies 

of seasonal variation of most terrestrial organisms 

correspond to the warmest period of the year from 

mid-spring to mid-autumn (Majadas & Urones, 2002), 

irrespective of latitude, elevation or habitat type (Bigot & 

Bodot, 1972; Hagvar et al., 1978; Paraschi, 1988; Lamotte 

& Blandin, 1989; Radea, 1993; Deltshev & Blagoev, 1994; 

Assi, 1998; Chatzaki, 1998; Chatzaki et al., 1998), the 

winter activity of spiders is also very important and it is 

mostly restricted to the families Linyphiidae, Dysderidae, 

Clubionidae and Lycosidae (Schaefer, 1977; Radea, 

1993; Chatzaki et al., 1998). 

Organisms are able to withstand low temperatures of 

the winter in several ways: 

a) by using physiological adaptations, such as 

reduction of metabolic rate to a negligible amount, 

or cryptobiosis (Barra et al., 1989), and lowering of 

the supercooling point, by accumulating antifreezing 

compounds (Block, 1996; Sømme, 1996), 

b) by using behavioral adaptations, such as winter 

diapause, restriction of activity, vertical migrations 

and search for winter retreats (Stamou, 1998), 

c) by synchronizing their biological cycle with the 

seasons and overwintering in the egg phase, which 

is the most tolerant developmental stage (Schaefer, 

1977; Aitchison, 1984). 

Spiders which are active during the dry period 

increase their hunting activity at night 1 , retreating to 

shady places (under stones, rocks or litter) during 

the day, or use vertical migrations to more favourable 

microenvironments (Duffey, 1968; Di Castri & Vitali, Di 

Castri, 1981; Cloudsley-Thompson, 1982; Lamotte & 

Blandin, 1989), in order to withstand the limiting factors 

of this period. Cryptobiosis and aestivation are also 

types of physiological adaptations to the summer stress, 

but these probably concern organisms which are more 

vulnerable to aridity than spiders (Stamou, 1998). 

Seasonality in the activity of spiders (Cloudsley- 

Thompson, 1978), as well as of most invertebrates 

(Pearson & White, 1964; Iatrou & Stamou, 1989; 

Lamotte & Blandin, 1989; Stamou, 1998), is mainly 

influenced by photoperiod and temperature, in relation 

to the physiological tolerance of individual species to 

temperature and humidity. These are the main factors 

controlling survival (Wise, 1993). However, experimental 

data has shown that prey densities (i.e. Diptera, Isopoda, 

Thysanura, Acarea, Collembola and Psocoptera) are 

important factors, which influence the abundance of 

spiders more than abiotic factors (i.e. air temperature, 

humidity and precipitation) in a pine forest on the island 

of Skopelos, Greece (Radea, 1993 and references therein). 

The latter affect spider densities only in an indirect way. 

The observation that the majority of species overwinter 

as immatures, and not as eggs, in order to be able to take 

advantage of the available prey early in the spring, shows 

that, despite the harshness of the climate during winter, 

the main determining factor of the biological cycles of 

spiders is food availability (Schaefer, 1977). For most 

arachnid species, the increase in availability of food and 

in daylight, enhance reproduction towards the end of 

spring and lead to an increase in their activity at that 

time (Coulson et al., 1975; Paraschi, 1988; Majadas & 

Urones, 2002). 

There are several patterns in the literature which 

describe the seasonal variation of spiders. According to 

Aitchison (1984) the following categorization is used as 

a standard: 

a) eurychronus species, having adults present 

during the whole or most of the year. This pattern was 

earlier defined as diplochronism by Tretzel (1954) to 

describe the double peak of activity in males, observed 

from pitfall traps collections. It is not clear when the 

reproductive period(s) of this pattern occurs (Toft, 

1. Nocturnal activity is common in most arachnids, especially those of the arid or semi-arid ecosystems. It is related to biotic (predation 

avoidance) and abiotic (reduction of water loss) factors (Cloudsley-Thompson, 1978). 


PHENOLOGICAL PATTERNS OF GROUND SPIDERS (ARANEAE) ON CRETE, GREECE ◆ 

1976; Milner, 1988) and in general there is more than 

one type of biological cycle that could explain it, i.e. 

half, one or two years (Toft, 1976). 

b) stenochronous species, having adults present 

at a certain time of the year that corresponds to the 

reproductive period. Depending on the season when 

this happens, the pattern may be further divided to 

spring, summer and autumn stenochronous (Schaefer, 

1977; Foelix, 1981). When pitfall traps are used, these 

patterns are represented by one peak of male activity, 

indicating a search for females and copulation, 

followed by (or almost at the same time) a peak in 

female activity, indicating the intense search for food 

(Tretzel, 1954) and/or for finding suitable places 

(Duffey, 1956) just prior to egg deposition. Usually, 

this is followed by an increase of immature stages that 

represent the following generation. These patterns may 

be attributed to annual or biennial cycles. In the latter 

case, mature individuals are observed for a longer 

period (from may to october) but each generation 

matures in two years, whilst reproduction takes place 

in a shorter period (june-july) (Toft, 1976). 

c) winter-active species which are mostly active and 

reproduce during winter months. 

Studies analyzing the phenology of spiders at species 

level in Mediterranean ecosystems are very scarce (Assi, 

1986; Majadas & Urones, 2002). Other studies have 

been carried out at family level (Bigot & Bodot, 1972; 

Christophe, 1974; Paraschi, 1988; Chatzaki et al., 1998) 

or refer to different bioclimatic zones (Tretzel, 1954; 

Toft, 1976; Aitchison, 1984; Milner, 1988; Deltshev & 

Blagoev, 1994). Although not shown earlier, it is unlikely 

that studies based on family level, albeit indicative of the 

actual phenological patterns of the taxa, are sufficient to 

account for the seasonal variation of ground spiders in 

Mediterranean ecosystems, since this is directly linked 

to the species’ biological cycles and to the climate of 

a region. In this study the phenology of one of the 

dominant families of ground spiders, Gnaphosidae, is 

presented. Phenological patterns of the 11 most common 

and abundant species of Crete, Greece are compared in 

relation to each other and among the 17 sites of the study 

area. The main goal of the present study is to answer the 

following questions: 

1. What is the phenological pattern followed by each 

one of the dominant species of the family Gnaphosidae 

on Crete? 

2. How stable is this pattern along the sites of Crete, 

knowing that there is great habitat and climate 

heterogeneity along this island? What are the main 

reasons for any possible variation? 

3. What is the phenological pattern of the family as 

a whole at each site and how informative can this 

pattern be? 


Study area 

Crete is the largest island of the south Aegean island 

arc and is situated at its centre (34°.50’-35°.40’N latitude, 

23°.30’-26°.20’E longitude). It is the fifth largest island 

in the Mediterranean, after Sicily, Sardinia, Cyprus 

and Corsica, with total area covering 8.261 km 2 . Crete 

presents great geomorphological variation, being a 

“miniature continent”, as Rackham & Moody (1996) 

very illustratively report. Even within such a small area, 

a great variety of habitats and climatic factors are present, 

ranging from the insular character of the coasts to a fully 

continental character in the centre, where three main 

mountains of high altitude (peaks of 2 000-2 400 m) 

exist: Lefka Ori in the western part, Psiloreitis in the 

central part and Lasithiotika Ori (Dikti and Thrypti) in 

the eastern part. As far as landscape, climate, vegetation 

and human impact are concerned, Crete is very 

asymmetrical along its main axes, north-south, west-east 

and along the altitudinal gradients. 

Climate. The climate of Crete is typical Mediterranean, 

with 5-7 months of arid, hot and dry summers, alternating 

with smaller periods of rainy, mild winters. Along the 

main axes of the island there is a remarkable gradient 

of temperature, precipitation, winds and humidity. The 

south, east and inland lowlands are warmer and drier than 

the rest of regions (Rackham & Moody, 1996). 

Mean annual temperatures are 2 °C higher in the south 

than in the north of Crete (Pennas, 1977) and they fall by 

6 °C per 1 000 m altitude (Grove et al., 1991). The range 

of temperatures is much narrower near the coasts than 

in the mountains, due to the more maritime character of 

the former (Strid, 1995). The southeastern coast of Crete 

is the warmest and most arid area throughout Greece 

(relative humidity can be as low as 20 % in the summer), 

while the northwestern parts of Lefka Ori belong to the 

more humid zones (Pennas, 1977; Strid, 1995). 

Mean annual precipitation for Crete is 400-650 mm. 

However, there is considerable variation in both time 

and space. The rainy period extends from october to 

35 



36 

march (mainly from december to february), when 85- 

90 % of the total rain falls (Pennas, 1977). Precipitation 

is higher at the western and northern parts of Crete, as 

well as at the mountains. Although there are no records 

about the precipitation above 900 m, Rackham & Moody 

(1996) estimate that at the top of Lefka Ori the annual 

precipitation must be as high as 2 000 mm, while at the 

southeastern corner of Crete as low as 240 mm. 

Vegetation. The dominant plant formations are 

phrygana and maquis, which on Crete are often 

intermixed due to overgrazing and human activities. 

Phrygana is the most widespread formation on the island. 

Lowland and middle altitude phrygana are composed of 

thorny aromatic shrubs, such as: Sarcopoterium spinosum, 

Coridothymus capitatus, Phlomis spp., Cistus spp., Genista 

acanthoclada, Calicotome villosa, Euphorbia spp., Ballota 

spp. and many others. At higher elevations other species 

(chamephytes) dominate, such as: Berberis cretica, 

Rhamnus saxatilis, Prunus prostrata, Satureja spinosa, 

Astragalus angustifolius, etc. The upper limits of forest 

growth reach altitudes from 1 600 m (southern slopes) to 

1 800 m (northern slopes). Pinus brutia is the dominant 

tree species up to 1 200 m. This is the commonest forest 

formation on Crete, being found from sea level to such 

elevations. Beyond the timberline, only scarce vegetation 

composed of small cushion-like shrubs exists. 

Sampling 

For the purposes of the present study 17 sites were 

selected along Crete, in order to cover its horizontal and 

vertical axes and the altitudinal gradients of the three 

main mountain massifs, Lefka Ori, Psiloreitis and Dikti 

(fig. 1). The vegetation type of most of the lowland sites 

is phrygana. Details for each site are given in table 1. 

At all sites material was collected using pitfall traps 

containing ethylene glycole as preservative. Pitfalls were 

set and changed at two-month intervals 2 . Collection 

of material lasted for one year at each site. Although 

this period does not sufficiently describe the temporal 

variation of species during an entire bioclimatic cycle 

– which in northern Greece (Stamou, 1998) as well as 

in most Mediterranean regions (Petanidou, 1991) lasts 

four years – it is however quite commonly used in similar 

analyses (Radea, 1993; Groppali et al., 1996; Assi, 1998). 

Fig. 1 Map of sites 

in the study area. 

2. In sites 3-6 and 11, material was collected every month. In most cases collecting periods were modified to shorter intervals in order to 

conform to the rest of data, unless the information they provided was very important, so they were presented in monthly intervals. 



Sites Locality Altitude (m) Habitat type Dates of collection and corresponding groups for analysis 

Site 1 Gramvousa 0 Phrygana at sea level. 

a: 25/4/1996-26/6/1996; b: 26/6/1996-23/8/1996; c: 23/8/1996- 

29/10/1996; d: 29/10/1996-30/12/1996; e: 30/12/1996-14/3/1997; f: 

15/3/1997-12/5/1997. (A) 

Site 2 Elafonisi 0 

Phrygana-maquis (Juniperus oxycedrus, Pistacia lentiscus, 

Coridothymus capitatus and Ceratonia siliqua). Highly 

disturbed area due to touristic activities. 

a: 25/4/1996-26/6/1996; b: 26/6/1996-25/8/1996; c: 25/8/1996- 

29/10/1996; d: 29/10/1996-30/12/1996; e: 30/12/1996-13/3/1997; f: 

13/3/1997-7/5/1997. (A) 

Site 3 

Lefka Ori 

(above Anopoli village) 

800 

Mature pine forest (Pinus brutia) with very little undergrowth 

consisting mainly of Quercus coccifera. 

a: 18/10/1990-23/11/1990; b: 1/3/1991-28/3/1991; c: 28/3/1991-5/ 

5/1991; d: 5/5/1991-8/6/1991; e: 8/6/1991-6/7/1991; f: 6/7/1991-4/ 

8/1991; g: 4/8/1991-8/9/1991; h: 8/9/1991-5/10/1991; i: 5/10/1991- 

6/11/1991; j: 6/11/1991-7/12/1991; k: 7/12/1991-11/1/1992; l: 11/1/ 

1992-8/3/1992; m: 9/3/1992-5/4/1992. (E) 

Site 4 

Site 5 

Site 6 

Lefka Ori 


Lefka Ori 


Lefka Ori 


1200 

1600 

2000 

Old cypress wood (Cupressus sempervirens) with scarce 

Quercus coccifera and Acer sempervirens. Traps of this period 

were considered active only during the first 30 days 

Plateau over the timberline. Scarce vegetation consisting of 

cushion-like bushes, mainly Juniperus oxycedrus oxycedrus, 

Berberis cretica, Prunus prostrata, Satureja spinosa. 

Valley with scarce vegetation consisting of Berberis cretica, 

Prunus prostrata, Astragalus angustifolius and Satureja spinosa. 

a: 18/10/1990-23/11/1990; b: 1/3/1991-28/3/1991; c: 28/3/1991-5/ 

5/1991; d: 5/5/1991-8/6/1991; e: 8/6/1991-6/7/1991; f: 6/7/1991-4/ 

8/1991; g: 4/8/1991-8/9/1991; h: 8/9/1991-5/10/1991; i: 5/10/1991- 

6/11/1991; j: 7/11/1991-4/5/1992 4 . (E) 

a: 29/7/1990-1/9/1990; b: 1/9/1990-17/10/1990; c: 18/10/1990-23/11/ 

1990; d: 28/3/1991-5/5/1991; e: 5/5/1991-8/6/1991; f: 8/6/1991-6/7/ 

1991; g: 6/7/1991-4/8/1991; h: 4/8/1991-7/9/1991; i: 7/9/1991-5/10/ 

1991; j: 5/10/1991-6/11/1991; k: 6/11/1991-6/6/1992 4 . (E) 

a: 29/7/1990-1/9/1990; b: 1/9/1990-16/10/1990; c: 8/6/1991-6/7/1991; 

d: 6/7/1991-4/8/1991; e: 4/8/1991-7/9/1991; f: 7/9/1991-6/10/1991; g: 

6/10/1991-7/8/1992. (E) 

37 

Site 7 Exantis 200 

Dense phrygana-maquis dominated by Quercus coccifera, 

Sarcopoterium spinosum, Corydothymus capitatus and 

Calicotome villosa. 

a: 22/4/2000-6/7/2000; b: 6/7/2000-14/9/2000; c: 14/9/2000-7/11/ 

2000; d: 7/11/2000-13/1/2001; e: 13/1/2001-12/3/2001; f: 12/3/2001- 

8/5/2001. (C) 

Site 8 

Psiloreitis 

(above Kouroutes village) 

1650 Subalpine shrubs. 

a: 14/4/2000-2/7/2000; b: 2/7/2000-14/9/2000; c: 14/9/2000-30/10/ 

2000; d: 30/10/2000-24/3/2001; e: 24/3/2001-12/6/2001. (D) 

Site 9 

Psiloreitis 

(above Lochria village) 

1950 Subalpine shrubs. 

a: 14/4/2000-2/7/2000; b: 2/7/2000-15/9/2000; c: 15/9/2000-30/10/ 

2000; d: 30/10/2000-24/3/2001; e: 24/3/2001-12/6/2001. (D) 

Site 10 Panagia Almyri 100 

Site 11 Youchtas 200 

Site 12 Keratokampos 0 

Phrygana (Calicotome villosa, Sarcopoterium spinosum, 

Phlomis sp.) and Nerium oleander by the riversides. 

Phrygana (Quercus coccifera, Genista acanthoclada, 

Corydothymus capitatus, Ebenus creticus and Salvia fruticosa). 

Phrygana-maquis close to the beach. Abandoned cultivations 

close to the village. 

a: 16/3/1999-20/5/1999; b: 20/5/1999-26/7/1999; c: 26/7/1999-30/9/ 

1999; d: 30/9/1999-26/1/2000; e: 26/1/2000-2/7/2000. (B) 

a: 16/12/1995-15/1/1996; b: 15/1/1996-18/2/1996; c: 18/2/1996-2/4/ 

1996; d: 2/4/1996-28/4/1996; e: 28/4/1996-6/6/1996; f: 6/6/1996-2/7/ 

1996; g: 2/7/1996-8/8/1996; h: 8/8/1996-16/9/1996; i: 16/9/1996-10/ 

10/1996; j: 10/10/1996-14/11/1996 k: 14/11/1996-11/12/1996. (A) 

a: 27/11/1998-26/1/1999; b: 26/3/1999-26/5/1999; c: 26/5/1999-28/7/ 

1999; d: 28/7/1999-28/9/1999; e: 28/9/1999-26/1/2000. (B) 

Site 13 Milatos 300 Degraded phrygana. 

a: 21/4/2000-12/7/2000; b: 12/7/2000-11/10/2000; c: 11/10/2000-23/ 

1/2001; d: 23/1/2001-9/3/2001; e: 9/3/2001-6/5/2001. (C) 

Site 14 Dikti-Limnakaro 1450 Phrygana over the plateau. 

a: 11/5/2000-5/8/2000; b: 5/8/2000-2/10/2000; c: 2/10/2000-9/1/2001; 

d: 9/1/2001-10/5/2001. (D) 

Site 15 Dikti-Limnakaro 1750 Subalpine shrubs. 

a: 12/5/2000-5/8/2000; b: 5/8/2000-2/10/2000; c: 2/10/2000-9/1/2001; 

d: 9/1/2001-10/5/2001. (D) 

Site 16 Bramiana 0 

Lake reservoir divided into a pond and a steppe-like area 

resulting from desiccation. 

a: 30/10/1998-4/1/1999; b: 4/1/1999-3/3/1999; c: 3/3/1999-4/5/1999; 

d: 4/5/1999-22/7/1999; e: 22/7/1999-22/9/1999; f: 22/9/1999-1/2/ 

2000. (B) 

Site 17 Chamaitoulo-Xirokampos 280 Phrygana. 

a: 28/5/2000-6/8/2000; b: 6/8/2000-12/10/2000; c: 12/10/2000-23/1/ 

2001; d: 23/1/2001-16/3/2001; e: 16/3/2001-6/5/2001. (C) 

Table 1. Description of the study sites. 

Symbols in brackets (A to E) indicate sites where sampling was performed at the same year 

and that belong to the same group for the statistical analyses. 



38 

Experimental evidence also suggests that in spiders there 

is an impressive stability of phenological patterns, even 

between years of great climatic divergence (Hagvar et al., 

1978; Polis, 1998). It is also reported that the activity of 

organisms in Mediterranean ecosystems seems to depend 

much more on the seasonal variations within one year 

than on the variation among years (Stamou, 1998). Based 

on this idea, data from one year’s sampling were used here 

to describe the phenological patterns of species. 

Mature specimens (males and females) of the 11 most 

common and abundant species of the entire study area 

were counted at each site. All counts were transformed to 

numbers of individuals per 100 trap/days (td) in order to 

equate the sampling effort at all sites. Immature specimens 

were not taken into account, because taxonomical 

identification at species level was not possible. Phenology 

of the family at each site was measured as the sum of the 

numbers of mature individuals per sample. 

Phenological patterns are presented in rabdograms. 

The most characteristic phenograms from the sites in 

which each species occurred in a relatively high number 

of individuals are shown in the figures. The symbol 

(-) is used to simplify comparison between sites, when 

sampling periods did not fully correspond to each 

other, but do not correspond to an “actual” sampling. 

Phenology is presented as relative counts (%) of the total 

activity during the sampling periods. For each species, 

accompanying ecological comments on its habitat 

preferences are added. 

Statistical analysis 

Two Factor ANOVA was used to test differences in 

phenological patterns of each species between sites of 

the study area. Analysis was carried out with SPSS 8 

(SPSS Inc., 1989-1997). Seasons and sites were taken 

into account as the two factors, but only the sites 

comparisons were used in the results. Sites were divided 

into five groups of 3-6 sites (from A to E, see table 1), 

in which sampling was performed in the same year. For 

each species, counts (f) of male and female individuals 

were tested both separately and in total. Logarithmic 

transformation [log(f+1)] was used, because of great 

deviation among the counts. In order to be sure that 

the observed statistical differences did not result from 

differences in the absolute numbers of individuals, rather 

than of the activity patterns of the species, the analysis 

was also carried out by transforming the numbers 

of individuals into relative counts [log(rf+1)]. This 

transformation was calculated by dividing the number of 

individuals of a species collected at each period by the 

total number of individuals (of the same species) found 

at each site. 

In order to test statistically the similarities of patterns 

observed among species of this study, a further Two 

Factor ANOVA was performed using pairs of species as 

the first factor and season as a second. In this case we 

were only interested in testing the interaction between 

season and species groups. In those cases where this 

interaction was not significant, patterns of the species 

involved were considered to fall into the same group of 

phenological pattern. The test was performed for each 

site group separately. Different pairs of species were 

tested and those that showed the greater percentage of 

no interaction among groups of sites were used for the 

creation of a model representing the activity patterns 

which may be followed by Gnaphosidae on Crete. 

Percentages (rf) were arcsine-transformed, according to 

the formula y=arcsin(√rf) (Sokal & Rolhf, 1995). 

RESULTS 

Descriptive analysis 

Phenological patterns of species 

Callilepis cretica (Roewer, 1928) 

This species is quite common on Crete, occurring 

from sea level to 1 950 m. However it becomes more 

abundant at the zone 1 200-1 600 m and prefers rather 

cool and humid environments. In most lowland sites, 

mature individuals are found from may to september, 

with peak of males in may-june (fig. 2). Numbers of 

female individuals were too low to show a clear pattern. 

In the sites at higher elevations, sites 4-5 and 8-9, the 

activity of C. cretica is shortened and is shifted towards 

the end of summer. 

Drassodes oreinos (Chatzaki, 2002) 

This species occurs only on the mountains of Crete 

above 1 200 m and it becomes one of the dominant 

species above 1 650 m. At this latter elevation it is the 

only representative of the genus. The peak of male activity 

is in august-september (fig. 3). The activity of females 

is extended for a longer period, especially at the highest 

elevations. A similar, but less stable pattern, is observed 

in another high altitude species, Gnaphosa bithynica 



Fig. 2. Phenology of Callilepis cretica 

on Crete (relative counts (%) of the total 

catch per site). 

39 

Fig. 3. Phenology of Drassodes oreinos 

on Crete (relative counts (%) of the total 

catch per site). 



40 

(Kulczynski, 1903), with female individuals being present 

throughout the favourable period (Chatzaki, 2003). 

Drassyllus praeficus (L. Koch, 1866) 

This species is common on Crete from sea level to 

1 650 m. It prefers maquis of middle altitude, mainly in 

central Crete. Mature specimens are present from midspring 

until mid-summer (fig. 4). Comparing the patterns 

at all sites, it may be conjectured that the peak of males 

is in april-may and that of females is in may. According 

to Urones, Jerardino & Barrientos (1995), D. praeficus 

has two generations per year, and this is also suggested 

by the data presented here, where females do not seem 

to present a stable pattern, thus indicating longer activity 

during the dry season. 

Haplodrassus creticus (Roewer, 1928) 

This species is the most common representative of the 

genus Haplodrassus on Crete having no special habitat 

preferences. It has a narrow activity period early in the 

spring with peaks of both sexes occurring at some time 

between january and march with the female peak a bit 

later than that of the males (fig. 5). 

Micaria coarctata (Lucas, 1846) 

This species is the most common representative of the 

genus Micaria on Crete, reaching altitudes of 1 750 m. 

However it is not very active at the sites where it occurs, 

therefore no phenograms are presented for it. It has a 

rather stable phenological pattern with short activity 

towards the end of spring and beginning of summer 

(fig. 6). 

Nomisia excerpta (O.P. Cambridge, 1872) 

This species is one of the commonest on Crete. 

Mature specimens are found from mid-spring to end of 

summer (fig. 6) with male peak in may-june and female 

peak one or two months later. 

Fig. 4. Phenology of Drassyllus praeficus 

on Crete (relative counts (%) 

of the total catch per site). 



41 

Fig. 5. Phenology of Haplodrassus creticus 



Pterotricha lentiginosa (C.L. Koch, 1837) 

This species is the dominant species on Crete as far 

as both frequency and abundance are concerned, being 

common in most habitat types and altitudes of Crete up 

to 1 950 m. P. lentiginosa is one of the few species which 

is active in winter time as well as in the summer (fig. 7). 

It is also the only species that shows two distinct peaks 

of male activity (diplochronism) at most lowland sites 

and one female peak in may-june. At higher elevations 

the pattern changes with restriction of activity to only the 

warmer period. At these sites a single peak of male activity 

is observed in august-september, while the peak of female 

activity is observed a little earlier. Its close relative, P. kochi 

(O.P. Cambridge, 1872), shows similar activity pattern in 

Lebanon (Assi, 1998). 

Trachyzelotes malkini (Platnick, 1984) 

This species is the most common of all its congeners on 

Crete, reaching altitudes of 1 450 m. Mature specimens 

are found from mid-spring until the end of summer with 



42 

Fig. 6. Phenology of Nomisia excerpta 



male peak usually in may-june and female peak at the 

same time or a bit later (fig. 8). At higher elevations (i.e. 

site 14) activity is shifted towards the end of summer. 

Zelotes caucasius (L. Koch, 1866) 

This species is one of the commonest on Crete. Similar 

to most of its congeners (for instance Z. labilis, Z. scrutatus, 

Z. tenuis) the phenological pattern of Z. caucasius presents 

high activity throughout the dry period, with male peak 

in may-june and female peak at the same period or just 

afterwards (fig. 9). At higher elevations the pattern does 

not change, but it is more restricted to the summer 

months. 

Zelotes creticus (Kulczyski, 1903) 

This is a common species of middle and higher 

elevations of western Crete, but does not occur in eastern 

Crete. Its distribution at the lowland sites shows that it 

is rather sensitive to high aridity (Chatzaki et al., 2003). 

Its phenological pattern resembles that of D. oreinos with 



43 

Fig. 7. Phenology of Pteroticha lentiginosa 



Fig. 8. Phenology of Trachyzelotes malkini 





44 

Fig. 9. Phenology of Zelotes caucasius 



male peak in august-september and female peak in June, 

but with more extended presence of females during 

almost the whole year (fig. 10). 

Zelotes subterraneus (C.L. Koch, 1833) 

This species is very common on Crete. Together with 

P. lentiginosa, they form the main bulk of Gnaphosidae 

present during winter time. The male peak is observed 

in september-october at the lowland sites and in augustseptember 

at the higher elevations (fig. 11), and female 

peak at the same time or a bit later. It is interesting to note 

that Aitchison (1984) classifies this species as summer 

stenochronous, which, as will be commented upon in 

the discussion, is due to latitudinal difference of the two 

study areas. 

There are not any winter-active Gnaphosidae, 

consequently there are no species that appear only in 

winter months. Instead, the few Gnaphosidae that are 

present during this period belong to the most cold-tolerant 

species that present their maximum activity within the dry 

season. These are: P. lentiginosa, Z. subterraneus, H. creticus 

and also Anagraphis pallens, Z. cf. ilotarum, Z. scrutatus, 



Fig. 10. Phenology of Zelotes creticus 



45 

Fig. 11. Phenology of Zelotes subterraneus 





46 

Z. tenuis, which are not presented here, due to lower 

numbers of catches. 

Total seasonal variation of Gnaphosidae 

In this section, the total phenological pattern of the 

family at each one of the sites is presented. In all sites 

maximum activity of the family is observed within the 

dry season of the year, irrespective of their geographical 

position and prevailing bioclimatic conditions. The main 

difference between sites is in the exact time and duration 

of the high activity period. In each case the pattern is 

characterized by the dominant species at the sites. Based 

on these differences, phenological patterns of Gnaphosidae 

may be divided into three groups: 

a) those that have one narrow peak during the first 

period i.e. end of spring – mid summer (sites 1 and 

10), which is characterized by the dominance of 

members of the genus Zelotes. 

b) those that have one peak which lasts longer 

within the favourable period than that of (a) (sites 

12, 13, 16). These sites are characterized by the codominance 

of many species. Thus, the long activity 

period is produced by the conjunction of the activity 

patterns of many species whose peaks occur at 

different times within the favourable period: Zelotes 

spp. are responsible for the spring-summer activity, 

while P. lentiginosa, Z. subterraneus and Z. creticus are 

mainly responsible for the activity recorded later in the 

summer and early autumn 3 . 

c) those that present two peaks of activity (sites 2, 

7, 11, 17), in which P. lentiginosa and Z. subterraneus 

dominate, the latter adding mainly to the autumn 

peak. 

At higher elevations some sites shown one relatively 

long peak of activity, while others shown two peaks of 

activity. However the two peaks, when present, do not 

result from the presence of eurychronous species as 

in the case of lowland sites of group (c), but instead 

they are produced by many species: Drassylus praeficus, 

Callilepis cretica, Haplodrassus creticus, Zelotes caucasius in 

mid-spring, Nomisia excerpta, Z. tenuis, Z. creticus, Z. labilis 

in summer and Drassodes oreinos, Pterotricha lentiginosa, 

Z. creticus, Z. subterraneus at the beginning of september. 

Statistical analysis 

Differences between sites, based on the seasonal 

variation of each species (absolute counts), are shown 

in table 2. For most species there are not any significant 

differences in the phenological pattern they present 

among sites of the same group. For all species, statistically 

significant differences are recorded at sites of higher 

elevations (group E). P. lentiginosa is the only species in 

which differences in the activity pattern are recorded at 

a group of lowland sites (group C). When relative counts 

are taken into consideration (table 3), no differences in 

the patterns of each species are recorded, except for 

females of P. lentiginosa and Z. subterraneus between 

sites of group E. Consequently, most of the differences 

recorded in table 2 are derived from differences in the 

absolute numbers of the species at the sites of the study 

area. 

Taking into consideration the patterns of activity 

presented in the previous section, species were grouped 

in pairs and further tested for statistically significant 

differences (table 4). Non-significant P-values for the 

interaction between species and seasons indicate cases in 

which pairs of species follow the same pattern of activity 

within groups of sites. Missing values exist in cases where 

there was only one species at a sites group. Several trials 

were performed in order to find those pairs of species 

which showed the highest similarity of patterns. In group V 

there is a greater number of significant interaction which 

implies that the activity pattern of Z. subterraneus is not 

close to that of P. lentiginosa, hence leaving the latter 

species as the main representative of activity type V. 

DISCUSSION 

Model of phenological patterns 

From mid-spring to mid-autumn, Gnaphosidae is 

the dominant family of Cretan arachnofauna. In winter 

time the activity of most gnaphosid species is restricted 

and the family is outnumbered by other families, such 

as Linyphiidae, Dysderidae, Clubionidae (Chatzaki et al., 

1998). Within the favourable period, there is a continuum 

3. Other species which are less active, but are still present in early spring or early autumn and add to this pattern, are: Drassyllus praeficus, 

Haplodrassus creticus, Haplodrassus dalmatensis, Trachyzelotes malkini, T. lyonneti (spring), Zelotes solstitialis (autumn). 



Species Group Male Female Total 

Drassodes oreinos 

D 0,316 0,007 0,018 

E 0,086 0,004 0,004 

A 0,751 0,428 0,609 

B 0,949 0,391 0,837 

Nomisia excerpta 

C 0,391 0,799 0,433 

D 0,354 0,189 0,122 

E 0,054 0,013 0,011 

A 0,169 0,135 0,085 

B 0,069 0,191 0,149 

Pterotricha lentiginosa C 0,013 0,153 0,018 

D 0,049 0,118 0,668 

E 0,485 0,119 0,036 

Zelotes creticus E 0,726 0,007 0,039 

A 0,655 0,326 0,435 

B 0,735 0,425 0,493 

Zelotes subterraneus 

C 0,819 0,166 0,486 

D 0,628 0,221 0,336 

E 0,514 0,006 0,056 

Table 2. P-values of two factor ANOVA 

test among groups of sites – absolute counts of species. 

Values in bold indicate the significant results, 

in which phenological patterns within groups 

of sites are not the same. Species for which 

no significant values were recorded, 

are not shown in the table. 

47 

Species Group Male Female Total 

Pterotricha lentiginosa 

Zelotes subterraneus 

A 0,871 0,823 0,865 

B 0,358 0,744 0,901 

C 0,523 0,836 0,843 

D 0,744 0,768 0,688 

E 0,702 0,026 0,376 

A 0,183 0,750 0,800 

B 0,482 0,695 0,636 

C 0,983 0,863 0,803 

D 0,705 0,937 0,903 

E 0,784 0,007 0,108 

Table 3. P-values of single factor ANOVA test among 

groups of sites – relative counts of species. Values in bold 

indicate the significant results, in which phenological 

patterns within groups of sites are not the same. Species 

for which no significant values were recorded, are not 

shown in the table. 

in the peaks of species activity, a phenomenon that has 

already been observed by other authors, both in the 

Mediterranean (Urones, Jerardino & Barrientos, 1995) 

and in other temperate ecosystems (Enders, 1976; 

Toft, 1976; Uetz, 1977). This makes the division of 

phenological patterns into separate groups very difficult, 

especially when taking into consideration that the existing 

models of categorisation have been created to describe 

phenological patterns of spider species in very different 

ecosystems and latitudes. 

The best way to describe phenological patterns of 

Gnaphosidae of Crete – and perhaps of all Mediterranean 

ecosystems – is by dividing the two main periods of high 

activity into one which includes patterns with peaks 

from mid-spring to early summer, and another which 

includes patterns with peaks at the end of summer or at 

the beginning of autumn. These two rough groups may be 

further divided to subgroups, based on statistical analysis. 

As a result, the following categories of phenological 

patterns are proposed (fig. 12): 

I) Stenochronous species, with peak activity at the end 

of spring and/or beginning of summer (i.e. H. creticus, 

D. praeficus and M. coarctata). 

II) Species with peak activity similar to (I), but with 

longer total period of activity towards summer months 

(i.e. N. excerpta and Trachyzelotes spp.) 



Fig. 12. Suggested model of possible 

patterns of activity followed 

by Gnaphosidae on Crete. 

48 

Species group Site group Male Female Total 

A 0,071 0,021 0,001 

Table 4. P-values of interaction 

test (species*season) among grouped 

sites. Values in bold indicate 

cases in which interaction within 

groups of sites is not significant. 

(*) indicates cases of sites groups in 

which only one species was present 

and consequently no interaction 

could be detected. 

I 

II 

III 

IV 

V 

H. creticus, D. praeficus, M. coarctata 

Nomisia excerpta, Trachyzelotes malkini 

Callilepis cretica, Zelotes caucasius 

D. oreinos, Z. creticus 

P. lentiginosa, Z. subterraneus 

B - 0,342 - 

C 0,338 0,323 0,135 

D 0,274 0,796 0,299 

E - 0,183 - 

A 0,599 0,001 0,068 

B 0,511 0,517 0,693 

C 0,991 0,736 0,673 

D 0,453 0,629 0,001 

E


III) Species active during the whole favourable period, 

with peak activity in mid-summer (i.e. C. cretica, 

Z. caucasius and most other zelotine species such as 

Z. labilis, Z. scrutattus, Z. tenuis). 

IV) Stenochronous species with peak activity at the 

end of summer and/or beginning of autumn (i.e. 

D. oreinos and Z. creticus). 

V) Clearly eurychronous species with mature 

individuals occurring throughout the year and with 

two peaks of male activity (i.e. P. lentiginosa and to a 

lesser extent Z. subterraneus, which might as well be 

included in type IV). 

Although the definition of phenological patterns for 

the rest of the gnaphosid species occurring on Crete, due 

to low abundance, is not very clear, the few individuals 

of each species were collected during the first period of 

high activity. Consequently, these species possibly belong 

to patterns (I) and (II) (Chatzaki, 2003). Gnaphosa 

bithynica, Anagraphis pallens and Zelotes cf. ilotarum 

possibly belong to pattern (V), Zelotes solstitialis to pattern 

(IV) and Drassodes lutescens to pattern (II), but they all 

present high flexibility in their activity patterns. 

According to Grime (1974; 1979) organisms pursue 

different strategies for survival, resulting from three factors: 

competition, environmental stress and disturbance. Based 

on these three factors and used in conjunction with their 

phenologies, Milner (1988) divided the spiders of Oxleas 

Wood, UK into the following groups: 

a) stress-tolerant species, to which eurychronous or 

winter active species belong. 

b) high competitors, to which stenochronous species 

belong, but with extension of their presence during 

all summer. 

c) pioneer species, to which extreme stenochronous 

species belong. 

Following the above categorization, most Gnaphosidae 

of Crete (groups III and IV of our model) belong to 

the “high competitors” group. These species are very 

efficient competitors compared to other predators (even 

other spiders), concentrating their energy expenditure to 

the most advantageous period. Species of group V are 

stress tolerant, and they are considered to be catholic 

feeders, preying on animals that are present all year 

round, or, finally, they are unable to outcompete other 

more vigorous species during the period of optimum 

conditions. Species of group I may be considered as 

pioneers. Milner (1988) suggested that these species 

are related to highly disturbed environments, where they 

show a pioneer character, although they may appear as 

eurychronous or stenochronous elsewhere. 

Concerning Mediterranean ecosystems, several 

models have been proposed for the survival strategies of 

organisms. Asikidis (1989) suggested that organisms in 

Mediterranean ecosystems have to adapt in response to 

three factors: maximum habitat heterogeneity, maximum 

habitat predictability and moderate H/T (favourable 

reproductive period/generation period). According to 

Stamou (1998), Mediterranean arthropods have to 

be able to respond quickly to the abrupt changes in 

temperature, and to keep low levels of energy expenditure 

when temperatures remain stable, in order to optimize 

their energy balance. The same author postulates that 

these features define two main survival strategies: a) 

the conformists, i.e. organisms which synchronize 

their demographic patterns to seasonal changes, being 

more vulnerable to low temperatures and having long 

hibernating or aestivating periods and b) the conservatists, 

i.e. organisms that are stress tolerant and are active during 

longer periods, keeping low and stable energy equilibria. 

The synchronization of demographic events to seasons 

of the year implies critical ranges of environmental 

parameters, outside of which, organisms cannot respond. 

As a result, conformist species are related to narrower 

niche factors which include habitat and food. 

On an annual scale, most Gnaphosidae follow the 

conformist strategy, being mostly active during the 

dry period, but they possess mechanisms which allow 

them to respond to environmental conditions with 

more adaptability. If a narrow food selection results 

in the restriction of temporal and spatial distribution 

(Stamou, 1998), then the opposite should also happen. 

Hence, the generalized feeding of these spiders should 

allow them wider distributions in both time and space. 

A small percentage of species follow the conservative 

strategy, which corresponds to an even wider spatiotemporal 

distribution. Such species are found in most 

habitats of Crete, at most elevations and compose the 

main Gnaphosidae fauna in winter time. The main 

representative of this group – if not the only one – is 

P. lentiginosa. 

There is a view that processes which enhance the 

resistance of terrestrial arthropods to cold and to 

desiccation are complementary (Sømme, 1995; 1996), 

and that in most cases desiccation resistant species may 

be pre-adapted to cold tolerance (Block, 1996). This 

could explain why many of the Gnaphosidae occurring 

on the high mountains of Crete are those which are best 

adapted to the high aridity of lowland habitats. These 

49 



50 

gnaphosid species comprise the dominant group among 

spider families, as well as among arthropods of the 

ground (Lymberakis, 2003). They belong to the most 

tolerant, conservative species that can also change their 

phenological patterns along altitudinal gradients in order 

to adapt to mountain conditions. 

Studies on a different taxonomical 

level give different results 

Until recently, the phenological pattern characterizing 

most Mediterranean arthropods (Di Castri & Vitali-Di 

Castri, 1981; Lamotte & Blandin, 1989), spiders included 

(Bigot & Bodot, 1972; Christophe, 1974; Chatzaki et 

al., 1998; Majadas & Urones, 2002) and especially 

Gnaphosidae (Assi, 1986; Paraschi, 1988; Chatzaki et al., 

1998), was the one that presents two peaks of activity: 

one at the end of spring and one at the beginning of 

autumn. Also in many cases, the activity of spiders was 

documented as being continuously high (Hagvar et al., 

1978; Di Castri & Vitali-Di Castri, 1981; Radea, 1993). 

However, as it has been clearly shown in this study, 

phenological patterns are species specific (David, 1995; 

Stamou, 1998). The eurychronous pattern mentioned in 

the literature is followed by a small number of dominant 

species in the study area, and the high activity of the 

order as a whole, results from the conjunction of peaks 

of activity of different species, these peaks occurring at 

different times within the year. 

In the present study this deviation of species 

phenological patterns is shown for some lowland sites, 

where many species co-dominate, as well as in mountain 

sites. In the latter sites, the compartmentalization of 

species activity is maximized so as to allow co-existence 

in an environment where food availability becomes 

restricted. 

Phenological patterns are also different when different 

methods of collecting are used. Apart from the fact that 

pitfall traps do not give an accurate estimation of the 

densities of species (Uetz & Unzicker, 1976; Adis, 

1979), they also lead to certain types of phenological 

patterns, which mostly correspond to those behavioural 

tactics which increase their activity on the ground (for 

instance searching for a mate) (Toft, 1976). Differences 

between phenological patterns recorded by pitfall traps 

and other quantitative methods have been reported for 

other arthropods in Mediterranean ecosystems as well 

(Magioris & Tsiourlis, 1992). Paraschi (1988) used 

different methods to study spider communities in two 

maquis ecosystems in Greece. She concluded that the 

main differences observed when different methods are 

used, are the following: 

1. differences in the seasonal variation and peaks of 

activity, 

2. differences in the ratio of mature/immature 

individuals per sampling effort, 

3. differences in the family composition 

The reason why adult Gnaphosidae are not sampled by 

quadrate method may be related to the mechanisms they 

use to avoid high levels of aridity, i.e. to their nocturnal 

activity and to their hiding in more protected retreats 

during the day. 

Biological cycles in relation 

to phenological patterns 

Pitfall traps do not give accurate information about 

biological cycles of organisms. Field observations and 

female dissections would give clues to specify them 

more accurately. Although phenological patterns of 

species are usually stable and are definitely related to 

specific demographic events, there are often several 

ways to interpret them. For instance, many species of 

the genus Zelotes present one peak of activity, but they 

remain mature for long periods, even in winter (personal 

observations). Based on the analysis of Toft (1976), this 

pattern may be interpreted as stenochronous (perhaps 

with a biennial cycle) or as eurychronous, supposing 

that mature individuals survive for longer periods, but 

are not caught in the traps, due to low activity during 

winter months. 

More than half of spider species in Denmark (Toft, 

1976), Sweden (Almquist, 1969) and Canada (Aitchison, 

1984) are reported to have biennial cycles. In Central 

Europe most of the species are annual (Schaefer, 1977). 

Grimm (1985) verified this for Gnaphosidae, with 

copulation occurring in spring, egg laying in summer, 

egg hatching in summer and autumn and overwintering 

as immatures. These observations are in accordance 

with the results presented in this study. According to the 

same author, some species of the genus Zelotes may have 

two generations per year, while biennial cycles are rare 

in Gnaphosidae, except for species from higher latitudes 

(for instance Scandinavia, Lohmander, 1942). Similarly, 

biennial cycles are reported by Schmoller (1970) for 

four Gnaphosidae species in the alpine tundras of 

Colorado. Apparently the duration of biological cycles 

is dependent on the mean temperatures of a region, 

and on the duration of the cold period (Almquist, 1969; 



Schaefer, 1977; Aitchison, 1984). It is therefore expected 

that in Mediterranean climates, biological cycles are either 

annual or with two generations per year, for groups of 

animals which, in higher latitudes, may have biennial 

cycles. In Spain, Urones, Jerardino & Barrientos (1995) 

reported annual cycles for most Gnaphosidae. However, 

it is possible that D. oreinos and Z. creticus (as well as 

G. bithynica, not presented here) may have biennial cycles, 

as suggested by the more extended presence of female 

activity within the year, without being accompanied by 

increased activity of males (Toft, 1976). 

The change in the duration of biological cycles is a 

phenomenon widely reported in the literature, and is 

related to either latitude or altitude (Schmoller, 1970; Toft, 

1976 and references therein; Schaefer, 1977; Aitchison, 

1984). Often this variation is considered as an adaptation 

of species which extend their range of distribution to 

geographical zones with a different climate (Coulson et 

al., 1975; Reiskind, 1981), even though changes of annual 

cycles to biennial have been reported in the same area 

during “bad” years (Kajak, 1967) or vice versa during 

“good” years (Dondale, 1961). This flexibility of species 

to change their developmental rates at any time shows 

that their presence in time and space results from the 

combined effect of their own physiological tolerance and 

the environment. As a result the definition of their niche 

is a highly dynamic measure and not a static one, as was 

also observed by Lymberakis (2003). 

Ground spiders (Hagvar, Ostbye & Melaen, 1978) 

and especially Gnaphosidae (Deltshev & Blagoev, 1994) 

are considered to present stable phenological patterns, 

even when biological rates change for reasons previously 

mentioned (Toft, 1976). Along the vertical axis of Crete 

there are cases in which the activity of species is restricted 

(i.e. M. coarctata, Z. caucasius, P. lentiginosa) or is shifted 1-2 

months towards summer and early autumn (i.e. C. cretica, 

H. creticus, N. excerpta, T. malkini). In our view, these are 

responses to the more extreme climatic conditions of sites 

at higher elevations, without necessarily implying that a 

change in the biological cycle of these species actually 

takes place. According to Lymberakis (2003), the period 

between the end of summer and beginning of autumn is 

the most stable, and hence the most predictable, as far as 

climate is concerned, so it is selected by some organisms 

as being the most appropriate for reproduction. 

However, when the two generations per year change 

into one, then a change in the phenology of the species 

becomes evident (Toft, 1976). In Crete, this may be the 

case for P. lentiginosa, having two generations per year at 

the lowland sites and one at sites above 800 m. It could 

also be that the species has two reproductive periods 

at the lowland sites, leading to only one generation in 

springtime, so that an annual cycle takes place there as 

well. Even in this case, developmental stages occur at 

different moments along the altitudinal gradient, each 

having different duration. Two cycles, one short in the 

summer (5 months) and one longer in winter (7 months) 

is also reported for its congener P. kochi (O.P. Cambridge, 

1872), which also presents a similar phenological pattern 

(Assi, 1986). 


This work is part of the PhD thesis of the first author, 

which was funded by the Alexander Onassis Public 

Benefit Foundation and by the University of Crete. We 

are grateful to the working team of the Natural History 

Museum of the University of Crete for the collection and 

sorting of the material and to Dr K. Thaler for his help in 

the identification of species. We are also grateful to John 

Murphy for linguistic revision of the manuscript. 

References 

ADIS, J., 1979. Problems of interpreting arthropod sampling with 

pitfall traps. Zoologische Anzeitlung 202: 177-184. 

AITCHISON, C.W., 1984. The phenology of winter-active spiders. 

J. Arachnol. 12: 249-271. 

ALMQUIST, S., 1969. Seasonal growth of some dune-living 

spiders. Oikos 20: 392-408. 

ASIKIDIS, M. D., 1989. Dynamics and activity of oribatids 

(Acarea: Cryptostigmata) in evergreen – sclerophilous 

plant formations (Chortiatis). PHD thesis, University of 

Thessaloniki, Thessaloniki (in Greek). 

ASSI, F., 1986. Recherches écologiques sur le peuplement des 

araignées d’une garrigue du Liban. PHD thesis, Université 

de Paris IV, Paris. 

ASSI, F., 1998. Effets de la désertification dans l’Anti-Liban 

sur les populations d’araignées. I. Mise au point sur 

les Gnaphosidae Gnaphosinae (Araneae). II. Étude des 

populations de Thomisidae et Philodromidae (Araneae). Rev. 

arachnol. 12: 135-145. 

BARRA, J.A., BELGNAOUI, S. & POINSOT-BALAGUER, N., 1989. 

Sécheresse, froid même stratégies, un exemple type : le 

collembole Folsomides angularis. Bull. Ecol. 20: 65-66. 

51 



52 

BIGOT, L. & BODOT, P., 1972. Contribution à l’étude biocœnotique 

de la garrigue à Quercus coccifera. I. Étude descriptive 

de l’habitat et de la faune des invértébrés inventoriés. Vie et 

Milieu 23: 23-43. 

BLOCK, W., 1996. Cold or drought – the lesser of two evils for 

terrestrial arthropods? Eur. J. Entomol. 93: 325-339. 

BREYMEYER, A., 1966. Relations between wandering spiders 

and other epigeic predatory Arthropoda. Ecol. Polska (A) 

14: 27-71. 

CHATZAKI, M., 1998. Systematics and phenology of ground 

living spiders of the island group Gavdos – Gavdopoula. 

Master Thesis, University of Crete, Irakleio (in Greek). 

CHATZAKI, M., 2003. Ground spiders of Crete (Araneae, 

Gnaphosidae): taxonomy, ecology and biogeography. PHD 

Thesis, University of Crete, Irakleio (in Greek). 

CHATZAKI, M., THALER, K. & MYLONAS, M., 2003. Ground 

spiders (Gnaphosidae; Araneae) of Crete and adjacent areas 

of Greece. Taxonomy and distribution. III: Zelotes and allied 

genera. Rev. suisse Zool. 110: 45-89. 

CHATZAKI, M., TRICHAS, A., MARKAKIS, G. & MYLONAS, M., 

1998. Seasonal activity of the ground spider fauna in a 

Mediterranean ecosystem (Mt. Youchtas, Crete, Greece). 

Proceedings of the 17th European Colloquium of Arachnology, 

Edinburgh 1997: 235-243. 

CHRISTOPHE, T., 1974. Etude écologique du peuplement 

d’araignées d’une litière de châtaigneraie (forêt de Mont- 

Morency, Val d’Oise). Publications du laboratoire de Zoologie, 

Ecole normale supérieure, Paris, n o 3. 

CLOUDSLEY-THOMPSON, J.L., 1978. Biological clocks in 

Arachnida. Bull. Brit. arachnol. Soc. 4: 184-190. 

CLOUDSLEY-THOMPSON, J.L., 1982. Desert adaptations in 

spiders. Scientific Review of Arid Zone Residuals 1: 1-14. 

COULSON, J. C., HOROBIN, J.C., BUTTERFIELD, J. & SMITH, R. J., 

1975. The maintenance of annual life-cycles in two species 

of Tipulidae (Diptera); a field study relating development, 

temperature and altitude. J. Anim. Ecol. 44: 215-233. 

DAVID, J. F., 1995. Seasonal abundance of millipedes in a 

Mediterranean oak forest (Southern France). Isr. J. Zool. 

41: 23-31. 

DELTSHEV, C. & BLAGOEV, G., 1994. Biotopical distribution and 

seasonal activity of model species of the family Gnaphosidae 

(Araneae) in Zemen gorge (SW Bulgaria). Arachnol. Mitt. 

7: 20-30. 

DI CASTRI, F. & VITALI-DI CASTRI, V., 1981. Soil fauna of 

mediterranean-climate regions. In: Di Castri, D. W. Goodall 

& R. L. Specht (eds), Ecosystems of the World, Elsevier, 11: 

445-478. 

DONDALE, C.D., 1961. Life histories of some common spiders 

from trees and shrubs in Nova Scotia. Can. J. Zool. 39: 

777-787. 

DONDALE, C. D., REDNER, J.H. & SEMPLE, R.B., 1972. Diel 

periodicities in meadow arthropods. Can. J. Zool. 50: 1155- 

1163. 

DUFFEY, E., 1956. Aerial dispersal in a known spider population. 

J. Anim. Ecol. 25: 85-111. 

DUFFEY, E., 1968. An ecological analysis of the spider fauna of 

sand dunes. J. Anim. Ecol. 37: 641-665. 

ENDERS, F., 1976. Size, food finding and Dyar’s constant. 

Environ. Entomol. 5: 1-10. 

FOELIX, R.F., 1981. Biology of Spiders. Harvard University Press, 

Cambridge (Massachusetts). 

GRIME, J.P., 1974. Vegetation classification by reference to 

strategies. Nature 250: 26-31. 

GRIME, J.P., 1979. Plant strategies and vegetation processes. John 

Wiley, London. 

GRIMM, U., 1985. Die Gnaphosidae Mitteleuropas (Arachnida, 

Araneae). Abhandlungen des Naturwissenschaftlichen Vereins in 

Hamburg (NF) 26: 1-317. 

GROPPALI, R., PRIANO, M. & PESARINI, C., 1996. Fenologia 

araneologica (Arachnida, Araneae) in una siepe mista della 

pianura padana centrale. Bolletino del’Istituto Entomologico “G. 

Grandi” del’Universita di Bologna 50: 113-125. 

GROVE, A.T., MOONEY, J. & RACKHAM, O., 1991. Crete and 

the South Aegean Islands: effects of changing climate on the 

environment. Robinson College, Oxford. 

HAGVAR, S., OSTBYE, E. & MELAEN, J., 1978. Pit-fall catches 

of surface-active arthropods in some mountain habitats at 

Finse, south Norway. II. General results at group level, with 

emphasis on Opiliones, Araneida, and Coleoptera. Norw. J. 

Entomol. 25: 195-205. 

IATROU, G. D. & STAMOU, G. P., 1989. Preliminary studies 

on certain macroarthropod groups of a Quercus coccifera 

formation (Mediterranean-type ecosystem) with special 

reference to the diplopod Glomeris balcanica. Pedobiol. 33: 

301-306. 

KAJAK, A., 1967. Productivity of some populations of web 

spiders. In: Petrusewicz (ed.), Secondary productivity of 

terrestrial ecosystems, Warszawa-Krakow: 807-820. 

KUENZLER, E. J., 1958. Niche relations of three species of 

Lycosid spiders. Ecology 39: 494-503. 

LAMOTTE, M. & BLANDIN, P., 1989. Originalité et diversité des 

écosystèmes méditerranéens terrestres. Biol. Gallo-hellen. 16: 

5-36. 

LOHMANDER, 1942. Südschwedische Spinnen. Meddelanden fran 

Göteborgs Musei Zoologiska Avdelning 98: 1-163. 

LYMBERAKIS, P., 2003. Altitudinal differentiation of the fauna of 

Lefka Ori, Crete. PHD Thesis, University of Athens, Athens 

(in Greek). 

MAGIORIS, S. N. & TSIOURLIS, G. M., 1992. The role of the 

shrubs Quercus coccifera in the community structure of soil 

arthropods in two insular ecosystems (Is. Naxos, Cyclades, 



Greece). Procceedings of the 6 th International Conference of 

Mediterranean Climate Ecosystems, 140-146. 

MAJADAS, A. & URONES, C., 2002. Communauté d’araignées 

des maquis méditerranéens de Cyticus oromediterraneus Rivas 

Mart. et al. Rev. arachnol., 14: 31-48. 

MILNER, J.E., 1988. Oxleas Wood: Observations on the spiders, 

their phenology and ecological strategies. The London 

Naturalist 67: 97-118. 

MITRAKOS, K., 1980. A theory for Mediterranean plant life. Acta 

Œcologica, Œcol. Plant. 1: 245-252. 

PARASCHI, L., 1988. Study of spiders in maquis ecosystems of 

southern Greece. PHD Thesis, University of Athens, Athens 

(in Greek). 

PEARSON, R.G. & WHITE, E., 1964. Factors contributing to 

the annual cycles of surface-active arthropods in moorland 

country. Entomologist’s month Magazine, 201-206. 

PENNAS, P., 1977. The climate of Crete. PHD Thesis, University 

of Thessaloniki, Thessaloniki (in Greek). 

PETANIDOU, T., 1991. Epiconiasis in phryganic ecosystems. PHD 

Thesis, University of Thessaloniki, Thessaloniki (in Greek). 

PIANKA, E. R., 1994. Evolutionary Ecology (5 th edition). Harper 

Collins College Publishers, NY. 

POLIS, G. A., 1998. Multifactor population limitation: variable 

spatial and temporal control of spiders on Gulf of California 

Islands. Ecology 79: 490-502. 

RACKHAM, O. & MOODY, J., 1996. The making of the Cretan 

landscape. Manchester University Press. 

RADEA, C., 1993. Environmental factors that influence the 

temporal variation of predatory macroarthropods in the 

organic horizon of a Mediterranean pine forest. Biol. Gallohellen. 

20: 249-258. 

REISKIND, J., 1981. Phenology of two species of Castianeira 

(Araneae: Clubionidae) in central Florida: evidence of their 

geographic origins. Florida Entomologist 64: 457-458. 

SCHAEFER, M., 1977. Winter ecology of spiders (Araneida). 

Zeitschrift für angewandte Zoologie 83: 113-134. 

SCHMOLLER, R., 1970. Life histories of alpine tundra Arachnida 

in Colorado. The American Midland Naturalist 83: 119-133. 

SOKAL R.R. AND ROHLF F.J., 1995. Biometry (3 rd edition). W.H. 

Freeman and Co, NY. 

SØMME, L., 1995. Invertebrates in hot and cold environments. 

Springer, Berlin. 

SØMME, L., 1996. Anhydrobiosis and cold tolerance in tardigrades. 

Eur. J. Entomol. 93: 349-357. 

STAMOU, G. P., 1998. Arthropods of Mediterranean-type ecosystems. 

Springer, Berlin, NY, London. 

STRID, A., 1995. Desertification in the White Mountains of 

Crete. A botanical study with special reference to the 

effects of grazing and wildfires. Environment Research 

Programme 1991-1994: Climatology and Natural Hazards: 

Desertification in the Mediterranean area. 

TOFT, S., 1976. Life histories of spiders in a Danish beech wood. 

Natura Jutlandica 19: 5-40. 

TRETZEL, E., 1954. Reife und Fortpflanzungszeit bei Spinnen. 

Morph. Ökol. Tiere, 42: 634-691. 

UETZ, G. W., 1977. Coexistence in a guild of wandering spiders. 

J. Anim. Ecol., 46: 531-541. 

UETZ, G. W. & UNZICKER, J. D., 1976. Pitfall trapping in 

ecological studies of wandering spiders. J. Arachnol. 3: 101- 

111. 

URONES, C., JERARDINO, M. & BARRIENTOS, J. A., 1995. Datos 

fenologicos de Gnaphosidae (Araneae) capturados con 

trampas de caida en Salamanca (Espana). Rev. arachnol. 11: 

47-63. 

WILLIAMS, G., 1962. Seasonal and diurnal activity of harvestmen 

and spiders in contrasted habitats. J. Anim. Ecol. 31: 21-42. 

WISE, D. H., 1993. Spiders in ecological webs. Cambridge 

University Press. 

53 


On the seed ecology of two life forms of Spergularia 

(Caryophyllaceae) 

Écologie de la germination chez deux types biologiques différents 

de Spergularia (Caryophyllaceae) 

L.M.M. Bidak 

Botany Department, Faculty of Science, Alexandria University, Egypt. 

E-mail 

Abstract 

Two species representing two different life forms of the genus 

Spergularia have been chosen for the present study: the short lived 

perennial S. media (L.) C. Presl with smooth, brownish and winged 

seeds and the annual form S. bocconii (Scheele) Graebner, 

with brown, tuberculate and wingless seeds. The present study aims 

to compare between the two species regarding seed germination and 

the effect of salinity and to evaluate the relation between seed size 

and wingedness and to demonstrate the adaptive value of wings. 

The study showed that seeds of the perennial species attained larger 

size, mass and seed production than those of the annual species. The 

study also revealed that wingedness in the perennial species may 

serve in dispersal by water. Susceptibility to salt stress is increased 

after the emergence of radicle and plumule. Growth rates were 

strongly affected by increasing salinity and inhibition was clearer 

in the annual species. Larger seed size and delayed germination of 

the perennial species was associated with increasing salinity. 

Key-words 

Seed germination, seed production, dispersal, salinity 

Résumé 

Deux espèces du genre Spergularia représentant deux types biologiques 

différents ont été choisies dans cette étude : S. media (L.) 

C. Presl est une espèce pérenne éphémère qui possède des graines 

brunâtres, lisses et ailées, et S. bocconii (Scheele) Graebner, espèce 

annuelle avec des graines aptères brunes et tuberculeuses. Cette 

étude a pour objectifs de comparer les deux espèces sur le plan de 

la germination des graines et de tester l’effet de la salinité. Elle 

vise également à évaluer la relation entre la taille des graines et la 

présence d’ailes ainsi que de démontrer la valeur adaptative de ces 

appendices ailés. Cette étude montre que l’espèce pérenne possède des 

graines de taille, poids et production plus importants que celle de 

l’espèce annuelle. L’étude montre également que la présence d’ailes 

chez l’espèce pérenne peut servir à la dispersion par l’eau. La sensibilité 

au sel s’accroît après l’émergence de la radicule et de la plumule. 

Les taux de croissance sont fortement affectés par un accroissement 

de la salinité et une inhibition apparaît clairement pour l’espèce 

annuelle. Le retard de germination des graines de grande taille de 

l’espèce pérenne est associé à un accroissement de la salinité. 

Mots-clés 

Germination des graines, production de graines, dispersion, 

salinité 

55 


◆ L.M.M. BIDAK 

56 

INTRODUCTION 

Well-developed dispersal structures are more common 

in perennial than in annual species (Werner, 1979). 

Venable and Levin (1983) argue that Werner’s conclusion 

supports the idea that dispersal in space may be more 

important for perennials than for annuals. Unfortunately 

both studies lack data on seed mass. Differences in seed 

size are always associated with differences in germination 

rate and establishment and variation in seed shape may 

lead to dissimilar dispersal of different seed types 

(Telenius & Torstensson, 1991). Seeds can become more 

complicated by adaptation to special modes of dispersion 

(hairs, wings, appendages, etc.). However, winged seeds 

are generally wind dispersed (i.e. anemochorous) (Rauh 

et al., 1975). 

The number of species in the study area varies from 

habitat to another. The total number of perennial species 

in the various habitats of the study area was 42. On the 

other hand, the number of annuals encountered was 52. 

About 27 % of the species are restricted to the narrow 

coastal strip (ca. 15 km broad) along the Mediterranean. 

About 19 % of the species are also present in this coastal 

strip, and extend their distribution to the Nile delta and 

the newly reclaimed lands (El-Ghareeb & Rezk, 1989). 

For most of the desert plants soil moisture and 

rainfall are limiting factors influencing seed germination 

and establishment (Freas & Kemp, 1983: Agami, 

1986). Harper et al. (1965) found that variation in the 

microtopography of the soil surface could determine 

the success or failure of establishment of seedlings. The 

shape and size of seeds may interact with heterogeneity 

in soil microtopography, which determines the abundance 

and proportions of different species in the community. 

One topic of germination ecology that needs further 

investigation is the pH requirements for germination 

of seeds under high salinity. Seeds of a number of 

species may have the ability to increase the hydrogen ion 

concentration in their immediate surroundings and thus 

lower the pH and hence may succeed to germinate under 

high salinity conditions (Okusanya, 1978). Germination 

inhibitors and salts may occur in the dispersal units (or 

in other parts of the plants) act as “rainclocks” which 

insure that no germination will take place until the soil 

is wetted by several rains and seeds of some species do 

not germinate until the salts leached out of the seed’s 

microhabitat (Chapin, 1991). 

Spergularia (Pers.) J. et C. Presl (Caryophyllaceae, 

subfamily Parnychioideae) is commonly found in saline 

areas and salt marshes. The genus is represented in 

Egypt by five species, mostly in two life-forms; annual 

or short-lived perennials (S. rubra, S. boconnii, S. diandra 

and S. marina) and perennial herbs (S. media), sometimes 

with a woody base (Boulos, 1995). Seeds are winged or 

wingless (Boulos, 1995). Many species (e.g. Spergularia 

rubra) have medicinal value. The medicinal action of the 

plant is due to the large proportion of aromatic resins. 

The genus Spergularia is rich in saponins, which is widely 

used as a domestic detergent (Neimann, 1993). 

The aims of this study is to provide informations 

about the seed germination of the two investigated 

species, a perennial (Spergularia media) and an annual 

(S. bocconii) and to compare the degree of adaptation for 

habitat performance and the relation between seed size 

and wings between the two species. 


Study area 

The study area extends for a distance of about 20 km 

along the Mediterranean coast from lake Edku in the east 

to Rosetta in the west and to a depth of 2 km. The climate 

is arid, warm coastal desert type (Meig, 1973), with a 

mean winter temperature above 10 °C and rainy during 

winter. The annual rainfall ranges from 91 to 175 mm. 

The mean relative humidity is lower in summer than in 

winter (65 % and 81 % respectively) and evaporation is 

higher in summer than in winter (7.8 and 2.8 mm/day 

respectively). Eight main physiographic categories were 

distinguished: 1- unstabilized dunes, lying close to the 

seashore; 2- stabilized dunes distinguished into leeward 

side, the windward side, and the summit; 3- fields 

abandoned owing to rise of underground saline water; 

4- salt marshes covered by sand deposits from adjacent 

dunes; 5- salt flats; 6- canal banks; 7- cultivated orchards; 

and 8- cultivated crop fields. The coastal salt marshes 

extending from the foot of the youngest dunes to the sea 

shore and the inland salt marshes bordering the irrigation 

canals are the main habitats from where the studied 

species have been collected. 

Sampling strategy 

Mature individuals of the two selected species 

(Spergularia media and S. bocconii) were collected 

from their natural habitats. Four stands were selected 

to represent the natural habitats in which the studied 

species are distributed. These stands differ as regards to 


ON THE SEED ECOLOGY OF TWO LIFE FORMS OF SPERGULARIA ◆ 

their distance from the sea and the commonly associated 

species but do not differ much in their soil characteristics. 

Two stands located near the coast (400 m) and the others 

were 1 km away from the coast and located in the inland 

salt marshes. Fresh specimens were pressed, dried and 

kept in the herbarium of the Botany Department, Faculty 

of Science, Alexandria University. 

A bulk of seeds was collected from many individuals 

of both plant species. Seeds were air dried at laboratory 

temperature for several days, placed in polyethylene bags 

and stored at about 5 °C. Five individuals from each stand 

were used to study the lengths of shoot, lateral branches 

and root. The number of capsules per branch, seeds per 

capsules and the number of branches per individual were 

used to estimate seed production per plant. Seeds were 

weighed to the nearest mg, and seed dimensions were 

measured by microscope using a standard micrometer. 

Three soil samples at two depths (0-5, 5-15 cm) 

were collected from each site of the study area for the 

determination of pH and soil salinity. The concentration 

of Na, K, and Mg was determined using an atomic 

absorption spectrophotometer. The mixed acid digestion 

method was used for elements determination (Allen et 

al., 1974). One Way-ANOVA test was applied for the soil 

characteristics. 

Only few species tolerate more than 2 % NaCl solution 

(Chapman, 1966; Zahran, 1975). The success or failure of 

seed germination in saline solution is taken as an indicator 

of seed establishment and salt tolerance. For this purpose, 

seven concentrations of NaCl solution (0.1, 0.2, 0.3, 0.5, 

1.0, 1.5 and 2 % w/v that equivalent to 17, 34, 51, 86, 

172, 259 and 344 mm respectively) were prepared. Petridishes 

with filter paper moistened with distilled water or 

NaCl solution were prepared to examine the percentage 

seed germination under the different levels of salinity. 

Duplicate dishes, each with 30 seeds were used for each 

treatment. Observations were made every 24 hours. 

Emergence of radicle and/or plumule was taken as the 

criterion of successful germination. To assess the effect 

of salinity on germination rate the lengths of radicle and 

plumule (mm) were recorded daily for 8 days from the 

beginning of the experiment. All the data were subjected 

to one-way analysis of variance. All statistical analysis 

applied using MINITAB 13.1 release-PC computer 

program (Minitab, 2000) 

Results 

Description of physical, chemical characteristic of 

the soil and associated species of the natural localities is 

illustrated in table 1. The soil of the study area is loamy 

sand and alkaline (table 1). The habit of growth varied 

from prostrate to ascending. Individuals of S. media were 

taller than those of S. bocconii, with deeper roots. The main 

branches were 5-6 and were markedly longer in S. media. 

Leaves of S. media were linear and fleshy whereas those 

of S. bocconii were narrowly linear. Leaves and internodes 

were longer in S. media than in S. bocconii (table 2a). 

Average seed mass varied from 6.4 to 7.2 µg and seed 

length from 5.2 to 8.1 µm in S. media and S. bocconii 

respectively. The number of seeds/capsule was markedly 

higher in S. media and the number of seeds/individual 

was 1.4-fold the corresponding number in S. bocconii 

(table 2b). 

Seeds of S. bocconii can germinate immediately 

after being collected, whereas those of S. media exhibit 

delayed germination and seem to need a dormancy 

period. Hundred percent germination in S. bocconii under 

salinity conditions was attained at 0.1 % (17 mm) Nacl 

and thereafter germination rate declined steadily to reach 

approximately 27 % at 1.5 % (259 mm) Nacl within the 

end of the experimental period (18 days). 

Up to 0.3 % (51 mm) Nacl, the percentage of 

germination in S. media was 100 % thereafter germination 

rate declined so that 4 out of the 30 seeds germinated in 

2 % (344 mm) Nacl at the end of experimental period. 

At this salinity level, germination was totally inhibited in 

S. bocconii (table 3). It should be noted that during the 

first four days 26 seeds of S. bocconii germinated whereas 

no seeds of S. media could germinate. 

The maximum radicle and plumule lengths of S. media 

(56 and 75 mm respectively) were attained using distilled 

water. With increasing salinity, the length of both radicle 

and plumule were significantly decreased (fig. 1a). 

At 1 % (172 mm) NaCl concentration the radicle’s 

length of germinated seeds was more affected than the 

plumule’s. However, at 1.5 % (259 mm) NaCl no plumule 

emergence was observed. 

Highest germination rates were obtained for seeds 

of Spergularia bocconii wetted with distilled water or 

dilute NaCl solution (0.1 % (17 mm), 0.2 % (34 mm). 

Concentration over 0.2 % delayed germination rate 

and retarded the germination percentage (table 3). No 

seed of S. bocconii germinated at 2 % (344 mm) NaCl 

solution. The maximum radicle and plumule lengths (53 

& 65 mm respectively) were obtained using distilled water 

(fig. 1b). Lengths of both radicle and plumule as well as 

germination efficiency of S. bocconii were significantly 

affected by increasing salinization. No plumule was 

observed at 0.5 % (86 mm) NaCl solution. The growth 

57 



Soil Character Value ± S. E. Associated species 

pH 8.61± 0.17 Alhagi graecorum Boiss 

E.C. (mmohs cm -1 ) 2.93 ± 1.0 Phragmites australis (Cav.) Trin. ex Steud 

Soil Texture (%) 

Sand 

Silt 

Clay 

81.7 ± 4.4 

17.7 ± 2.9 

0.57 ± 0.46 

Juncus rigidus Desf. 

*Aeluropus lagopoides (L.) Trin. ex Thwaites 

Atriplex halimus L. 

Suaeda pruinosa Lange 

Soluble ions (meq./L.) 

58 

Table 1. Physical and chemical characteristics 

of the soil in the study area, the common associated 

species in the sites of the perennial (*) 

and that of the annual. 

Ca +2 

3.4 ± 1.8 

Mg +2 

Na + 

K + 

Cl - 

HCO 

- 3 

SO 

-2 4 

CaCO 3 

3.29 ± 1.17 

5.49 ± 2.67 

0.42 ± 0.13 

15.73 ± 4.62 

2.1 ± 0.49 

11.42 ± 3.24 

1.35 ± 0.29 

*Aster squamatus (Spreng.) Hieron. 

*Halocnemum strobiceum (Pall.) M. Bieb 

Inula crithmoides L. 

*Traganum nudatum Delile 

*Zygophyllum album L 

Juncus acutus L. 

*Silene succulenta Forssk. 

*Cakile maritima Scop. 

Organic matter (%) 0.17 ± 0.01 *Mesembryanthemum crystallinum L. 

Soil Type 

Loamy sand 

rate (mm/day) of both radicles and plumules of the 

two studied species decreased with increasing salinity. 

However, the growth rate of plumule exceeded that of 

radicle with a few exceptions (fig. 1a&b). 

DISCUSSION 

Telenius and Torstensson (1991) report that the 

habitat characteristics can explain the variability in seed 

size of the genus Spergularia. Presence or absence of 

wings, for example, affect seed dispersal trait and this 

pattern is tightly linked with seed size. In this investigation 

the seeds of the perennial species (S. media) attained a 

larger size and mass than those of the annual (S. bocconii). 

Differences in seed size between life histories have been 

reported by many authors (e.g. Baker, 1972; Pitelka, 

1977; Silvertown, 1981). 

Assuming that the pattern of seed dispersal is adaptive 

(Thompson, 1973; Augspurger & Franson, 1987; Matlack, 

1987), the evolution of larger seeds is constrained by its 

effect on seed dispersal. Thompson and Robinwitz (1989) 

hypothesized that big plants can produce big seeds and 

automatically brings enhanced seed dispersal. Venable 

and Levin (1983) found a correlation of wind dispersal 

adaptation and the perennial habit: well-developed 

dispersal structures were more common in perennial 

than in annual species. Telenius and Torstensson (1991) 

reported that the winged strategy is associated entirely 

with a perennial life history and reflected the fact that the 

perennial species of Spergularia produce larger seeds than 

do the annuals (fig. 2). 

Many authors emphasized the relation between 

the morphology of the diasporas and the dispersal 

agent (e.g. Rauh et al., 1975; Werner, 1979; Venable & 

Levin, 1983; El-Sheikh, 1996). Dansereau and Lens 

(1957) have previously proposed a simple classification 



A 

B 

59 

Figure 1. Growth 

rate (mm/day) 

of radicle (R) & 

plumule lengths (P) of 

S. media (fig. 1a) and 

S. bocconii (fig. 1b) 

after 2, 4, 6 and 8 

days from sowing at 

different concentration 

of salinity level and 

distilled water (C). 



Attribute Spergularia media Spergularia bocconii P Value 

Habit of growth Prostrate or ascending Ascending -- 

Plant height (cm) 53.4 ±2 21.3 ±0.9 *4.69 

Table 2a. Mean (at least 5 

individuals) ± SE of some 

morphological characters 

of the studied species. 

* P


Figure 2. Seed morphology 

and seed coat ornamentation 

pattern of S. media (A, B, C) 

and S. bobbonii (D, E) 

as shown by the SEM. 

61 

the conditions experienced by seeds in natural habitats 

depends on the possession of some information on the 

level of variability occurring between populations of 

different species occupying similar habitats, or with 

similar geographical distribution. 

In discussing the correlation between seed size and 

adaptive value of wings, Sterk (1996) focused on the seed 

size variation and dispersal characteristics in S. marina 

and S. media. He reported that the unwinged and winged 

seed types would represent respectively the r and K- 

strategies. Willson (1983) instead considered seed types 

as representative of different dispersal strategies. The 

present study may reveal that Spergularia species, which 

primarily inhabit “marshy” habitats, are present now in 

a range of habitats mainly as a result of human impact. 

Wingedness in the perennial species may serve in the 

pattern of secondary dispersal by water. 

Zayed and Zeid (1997) reported that there is an 

inverse relationship between the osmotic potential of 

the growth medium and the growth parameters. The 

present study showed that the susceptibility to salt stress 

is increased after the emergence of radicle and plumule, 

so that their lengths as well as their growth rates were 

strongly affected. The inhibition was notable in the annual 

than in the perennial species. The radicle length was more 

affected than the plumule length. This inhibitory effect 

is in agreement with the results obtained by Hajar et al. 

(1996) on their studies on Nigella sativa. 

Tolerance mechanisms require energy expenditure and 

may consequently be associated with large seed reserves 



Days 

Salinity Level (%) 

Distilled water 0.1 (17 mm) 0.2 (34 mm) 0.3 (51 mm) 0.5 (86 mm) 1.0 (172 mm) 1.5 (259 mm) 2.0 (344 mm) 

Spergularia bocconii 

62 

21/6/2002 10 9 1 0 0 0 0 0 

23 26 15 4 1 0 0 0 0 

25 30 24 12 5 11 3 2 0 

27 30 28 23 10 13 9 5 0 

29 30 28 23 22 18 10 8 0 

30 30 30 28 22 21 10 8 0 

1/7/2002 30 30 28 24 21 11 8 0 

3 30 30 28 25 23 11 8 0 

5 30 30 28 25 23 11 8 0 

6 30 30 28 25 23 11 8 0 

Spergularia media 

21/6/2002 0 11 5 7 8 0 0 0 

22 0 13 19 10 10 4 0 0 

23 0 19 23 19 18 4 0 0 

25 5 22 24 24 25 9 4 0 

27 12 24 24 26 26 13 6 0 

29 20 27 26 30 26 15 7 2 

30 20 27 26 30 27 19 7 2 

1/7/2002 22 27 28 30 27 19 7 2 

3 25 27 30 30 27 19 8 3 

5 25 27 30 30 27 19 8 4 

6 25 27 30 30 27 19 8 4 

Table 3. Results of germination percentages on the seeds of the studied species under different levels of salinity, 

compared with that of distilled water. Values are means of duplicate Petri dishes each containing 30 seeds. 

and slow relative growth rate (Nieman et al., 1992). 

Larger seed size of Spergularia media was associated 

with delay of germination under high salinity levels; this 

may enable seedlings to persist for longer time at low soil 

moisture and/or high soil salinity which might increase the 

chance of survival of these seeds. However, knowledge of 

the phylogenetic interrelations in the genus still requires 

further investigations. 


The authors appreciates the valuable comments of Dr. 

S. Barakat Prof. of Plant Physiology, Faculty of Science, 

Alexandria University Egypt. 



References 

AGAMI, M. 1986. The effect of different soil water potentials, 

temperature and salinity on germination of seeds of the 

desert shrub Zygophyllum dumosum. Physiol. Plant. 67: 305- 

309. 

ALLEN, S., GRIMSHAY, H.M., PARKINSON, J.A. & QUARMBY, C. 

1974. Analysis of Ecological Materials. Blackwell Scientific 

publications. Osney, Oxford, London, 565 p. 

AUGSPURGER, C.K. & FRANSON, S.E. 1987. Wind dispersal 

of artificial fruits varying in mass, area and morphology. 

Ecology, 68: 27-42. 

BAKER, H.G. 1972. Seed weight in relation to environmental 

conditions in California. Ecology, 53: 997-1010. 

BOULOS, L. 1995. Flora of Egypt Checklist. Cairo: Al-Hadara 

Publishing. 287 p. 

CHAPMAN, V.J. 1966. Vegetation and salinity. In: Salinity and 

aridity. Dr.W.Junk. The Hage, pp. 23-42. 

CHAPIN, F.S.III. 1991. Integrated responses of plants to stress. 

BioScience, 41: 29-36. 

DANSEREAU, P. & LEMS, K. 1957. The grading of dispersal 

types in plant communities and their ecological significance. 

Contrib. Inst. Bot. Univ. Montreal, 71, 52 p. 

EL-GHAREEB, R. & REZK, M. 1989. A preliminary study on 

the vegetation of the mediterranean coastal land at Bousseli 

(Egypt). J. Univ. Kuwait (Sci.), 16: 115-128. 

FREAS, K.E., & P.R. KEMP. 1983. Some relationships between 

environmental reliability and seed dormancy in desert annual 

plants. J. Ecol., 71: 211-217. 

HAJAR, A.S., ZIEDAN, M.A., AL-ZAHRANI, H.S. 1996. Effect of 

salinity stress on the germination, growth and some physiological 

activities of black cumin (Nigella sativa L.). Arabian 

Gulf J. Sci. Res., 14: 445-454. 

HARPER, J.L. WILIAMS, J.T. AND SAGAR, G.R. 1965. The behaviour 

of seeds in the soil. J. Ecol., 53: 273-86. 

MATLACK, G.R. 1987. Diaspore size, shape, and fall behavior in 

wind-dispersal plant species. Am. J. Bot. 74: 1150-1160. 

MCMILLAN, C. 1959. The role of ecotype variation in the distribution 

of the central grassland of North America. Ecol. 

Monogr., 29: 285-208. 

MEIG, P. 1973. World distribution of coastal deserts. In: Amiran, 

D.H.K. & Wilson, A. W. (eds.), Coastal deserts, their natural 

and human environments, Univ. of Arizona Press, Phoenix 

AZ, pp. 3-13. 

NIEMANN, G.J. 1993. Anthranilamide phytoalexins of the 

Caryophyllaceae. Phytochem., 34: 319. 

OKUSANYA,O.T. 1978. The effect of acid on the germination 

and early growth of some maritime cliff species. Oikos, 30: 

549-554. 

PITELKA, L.F. 1977. Energy allocation in annual and perennial 

lupines (Lupinus, Leguminosae). Ecology, 58: 1055-1065. 

RAUH, W., BARTHLOTT, W. & EHLER, N. 1975. Morphologie und 

Funktion der Testa staubformiger Flugsamen. Bot. Jahr. Syst. 

96: 353-374. 

SILVERTOWN, J.W. 1981. Seed size, life spain, and germination 

date as co-adapted features of plant life history. Am. Nat., 

118: 860-864. 

STERK, A.A. 1996. Biosystematic studies in Spergularia media 

and S. marina in the Netherlands III. The variatiability of 

S. media and S. marina in relation to the environment. Acta 

Bot. Neerl., 18: 561-577. 

TELENIUS, A. & TORSTENSSON, P. 1991. Seed wings in relation to 

seed size in the genus Spergularia. Oikos, 61: 216-222. 

THOMPSON, P.A. 1973. Seed germination in relation to ecological 

and geographical distribution. In: Taxonomy and Ecology, 

Heywood, V.H. (ed.), Academic Press, London, pp. 93-119. 

THOMPSON, K. & ROBINOWITZ, D. 1989. Do big plants have big 

seeds? Am. Nat., 133: 722-728. 

VENABLE, D.I. & LEVIN, D.A. 1983. Morphological dispersal 

structures in relation to growth habit in the Compositae. 

Pl. Syst. Evol., 143: 1-16. 

WERNER, P.A. 1979. Competition and coexistence of similar 

species. In: Solbrig, O.T., Jain, S., Johnson, G. & Raven, P.H. 

(eds.). Topics in plant population biology. Columbia Univ. 

Press. New York. 

WILLSON, M.F. 1983. Plant reproductive ecology. Wiley, New 

York. 

ZAHRAN, M.A. 1975. On the germination of seeds of Juncus 

rigidus C.A. Mey and J. acutus L. Bull. Fac. Sc. Mansoura 

Univ., 3: 75-84. 

ZAYED M.A., ZEID, I.M. 1997. Effect of water and salt stresses 

on growth, chlorophyll, mineral ions and organic solutes 

contents and enzymes activity in mung bean seedlings. Biol. 

Plant., 40: 351-356. 

63 


Factors structuring land snail communities in South-Eastern 

France: a comparison of two estimation methods 

Facteurs structurant les communautés malacologiques terrestres 

du SE de la France : comparaison de deux méthodes d’échantillonnage 

Sébastien Aubry* & Frédéric Magnin 

Institut méditerranéen d’écologie et de paléoécologie (IMEP-UMR 6116 CNRS), bâtiment Villemin, 

Europole de l’Arbois, BP80, F13545 Aix-en-Provence, CEDEX 04 

* Corresponding author: email: sebastien.aubry@univ.u-3mrs 

Abstract 

In order to shed light on the influence of sampling methods on the 

accuracy with which factors structuring land snail communities 

are gauged, the results of two studies were compared. The two 

studies, undertaken in the same area in South-Eastern France 

near Marseilles, differed in methods and initial aims. One was 

undertaken using a stratified sampling strategy which collected only 

large species within a 25 m 2 square while the other was undertaken 

using a systematic sampling strategy and collecting species of all sizes 

within a 400 m 2 area. 

Results indicate that thirty samples (using either technique) are 

sufficient to describe the main factors structuring land snail 

assemblages. In addition, the collection of only large species 

provides an adequate indication of the main factors structuring 

these assemblages. It is suggested that malacological quadrats are 

optimised when they cover a limited area (here 25 m 2 ) and include 

the collection of several turves. 

The comparison of the results of both studies revealed that human 

activities are the main factor structuring molluscan communities 

at the scale of the parish. It also highlights that this scale, which 

takes into account topography and human activities, is relevant to 

understanding the distribution of land snails. 

Key-words 

Community ecology, landscape, land snail, methodology, South- 

Eastern France, succession 

Résumé 

Dans le but de révéler l’influence de la méthode d’échantillonnage 

sur la précision avec laquelle les facteurs structurant les 

communautés malacologiques sont estimés, les résultats de deux 

études sont comparés. Ces deux études, entreprises dans la même 

région du sud-est de la France, près de Marseille, diffèrent par 

leurs méthodes ainsi que par leurs objectifs initiaux. L’une utilise 

un échantillonnage stratifié et ne prend en compte que les grosses 

espèces récoltées à l’intérieur d’un carré de 25 m² ; l’autre est fondée 

sur un échantillonnage systématique et prend en compte les espèces 

de toutes tailles à l’intérieur d’un carré de 400 m². 

Les résultats indiquent que trente échantillons (quelle que soit la 

technique) sont suffisants pour décrire les facteurs structurant les 

assemblages malacologiques à l’échelle d’une commune. De plus, 

la récolte des seules grosses espèces fournit une indication adéquate 

des principaux facteurs structurant ces assemblages. Il est suggéré 

que les quadrats malacologiques sont optimisés quand ils couvrent 

une surface limitée (ici 25 m²) et incluent la collecte de plusieurs 

prélèvements de sol pour la recherche des petites espèces. 

La comparaison des résultats des deux études révèle que les activités 

humaines sont le principal facteur structurant les communautés 

malacologiques à l’échelle de la commune. Elle souligne aussi que 

cette échelle, qui prend en compte la topographie et les activités 

humaines, est adéquate pour comprendre la distribution des 

mollusques terrestres au sein du paysage. 

Mots-clés 

Écologie des communautés, paysage, mollusques terrestres, méthodologie, 

Sud-Est de la France, succession 

65 


◆ S. AUBRY & F. MAGNIN 

66 

INTRODUCTION 

The role of micro-habitat in the composition of land 

snail communities has long been recognised (Boycott, 

1934). In particular, the structure of the vegetation is one 

of the main factors controlling the occurrence of species. 

The role of habitat heterogeneity on the structure of land 

snail communities has been studied at several scales and 

has been shown to be influential from the micro-scale 

(Davies & Grimes, 1999) to the landscape scale (Magnin 

et al., 1995). 

In contrast with phytosociology, where a standardised 

method exists, sampling methods have always been a 

source of debate in malacology and different methods 

have been used depending on the aims of the study (e.g. 

Cameron & Morgan-Huws, 1975; Bishop, 1977; Boag, 

1982; Emberton et al., 1996; Hawkins et al., 1998; Davies 

& Grimes, 1999; Menez, 2001; Cameron, 2002). 

The aim of the present work was to compare two 

studies in order to assess both the influence of sampling 

methods and strategies in general, and the factors that 

structure land snail communities in rural Mediterranean 

environments. The two studies sampled the same area of 

the French Mediterranean region but differed in their 

aims, sampling strategies and scales. One was conducted 

in 1989 (Magnin & Tatoni, 1995) and the second, the 

results of which are presented here, was conducted in 

1995. 

METHODS 

Study site 

Auriol is situated near Marseilles (coastal) in the 

calcareous lower Provence on the bank of the river 

Huveaune (fig. 1). It has a Mediterranean climate with 

a mean annual temperature of 14 °C, two dry months 

(P < 2T, where P is precipitation of a given month in 

mm and T is mean temperature of the month in °C) and 

two to three cold months (T < 7 °C) (CNRS, 1975). 

The parish of Auriol, delimited to the north and south 

by two mountain chains reaching 900 m, covers an area 

of 4 464 ha. Once an agricultural village (cultivated in 

terraces since at least the 12 th century), it is close to 

an enlarging urban centre, so that the landscape has 

undergone many changes since the beginning of the 

20 th century. Indeed, at the demographic maximum 

during the 19 th century, the entire area was covered with 

cultivated terraces. After World War II, abandonment 

of the agricultural terraces has led to the replacement 

of agricultural lands by woodland (Grove & Rackham, 

2001). More recently, the expansion of human habitations 

and roads has been influential. 

Aims of each method 

The 1989 study aimed to define post-agricultural 

successions on abandoned cultivation terraces in 

calcareous Provence for both plant and land snail 

communities (Tatoni et al., 1994; Magnin & Tatoni, 

1995). The 1995 study aimed to shed light on the many 

factors influencing land snail species diversity at several 

spatial scales within the parish of Auriol. 

Description of each method 

In 1989, all living and dead snails larger than five 

millimetres were taken during a standard period of fifteen 

minutes for each sampling site. This involved searching 

beneath fallen logs and stones, investigating crevices, 

under the bark of trees and in other places deemed 

suitable for snails. In total, 65 samples were collected 

along nine terraced hill slopes (fig. 1a). For each of the 9 

slopes several cultivation terraces were sampled according 

to their time of abandonment. Each sample was taken in 

an area of, approximately, 25 m 2 . 

In 1995, samples were taken according to a systematic 

sampling strategy whereby sampling sites were chosen a 

priori on a map. Twenty-five samples were taken from 

sites (20 x 20 m square) aligned on a grid (fig. 1b). 

Sampling was performed in two ways. First, a visual 

search involving the collection of all living and dead 

snails was undertaken for 20 minutes over the entire 

400 m 2 area. This involved searching beneath fallen logs 

and stones, investigating crevices, under the bark of trees 

and in other places deemed suitable for snails. Second, in 

order to collect shells less than 5 mm diameter, vegetation, 

litter and surface soil covering an area of 25 x 25 cm and 

a depth of 5 cm were collected at five points within each 

sampling site, bagged, and brought back to the laboratory. 

There, samples were dried in an oven (40-75 °C), then 

immersed in water. Floating material was collected in a 0.5 

mm mesh sieve and dried again. The sieved samples were 

separated into four fractions using a set of graded soil 

sieves (10, 2, 1 and 0.5 mm). Shells were then separated 

from plant material, using a binocular microscope for the 

smallest fractions. 


STRUCTURING FACTORS OF LAND SNAIL COMMUNITIES IN SOUTH-EASTERN FRANCE ◆ 

67 

Figure 1. Locations of sampling 

sites in the Auriol parish: a) 

1989; b) 1995. 



68 

ANALYSES 

Inventories of species collected during each study were 

compared. Nomenclature follows Kerney et al. (1983) 

except for Xeropicta derbentina (Krynicki, 1836), a species 

recently introduced in Provence that was not included by 

these authors. Randomised species accumulation curves 

(sample-based rarefaction curves) were calculated using 

the Estimate v6.0b1 software (Colwell, 2000) in order to 

assess the completeness of each inventory. Five hundred 

iterations were performed for each set of data. Species/ 

samples data matrices for each study were examined by 

correspondence analyses performed with CANOCO 

for Windows 4.0 (ter Braak & Smilauer, 1998). Species 

abundance data were logarithmically transformed (log 

(X +1)) in order to reduce the importance of the most 

abundant species and to normalise their distributions 

(Legendre & Legendre, 1984; Labaune & Magnin, 

2001). Also, rare species abundance was down-weighted 

by an algorithm available in CANOCO. 

RESULTS 

Inventory of biodiversity 

Thirty-two species were recovered from the 65 

samples collected in 1989 (table 1). Of these, four are 

small species (less than 5 mm) that were not specifically 

sought. The list of species recovered by this method 

therefore comprises only 28 species. 

Fifty-eight species were recovered from the 25 

samples collected in 1995 (table 1). Of these, five species, 

all hygrophilous, were represented only as sub-fossils, 

witnesses of a past environment. Therefore 53 species 

are found living in the area. Only 22 species were found 

during the visual search in 1995 (table 1). 

Many species, with various ecologies, were found 

during both studies. All the species found in 1989 were 

also collected during the 1995 study (table 1). Three 

categories of species missing from the 1989 study in 

comparison to the 1995 one can be highlighted. In fact, 

in addition to the expected absence of species less than 

5 mm in size, open-wetland species and forest species 

were also missing. This absence of species from the latter 

categories is a result of the 1989 sampling strategy that 

deliberately ignored habitats such as irrigated agricultural 

crops and closed woodlands that were not present on 

abandoned terraces. Thus, despite a much larger number 

of samples, some large species were not recovered from 

the 1989 study. 

Accumulation curves 

The method used in 1989 led to collection of a 

large number of species bigger than 5 mm. The sample 

accumulation curves (fig. 2), computed using EstimateS 

(Colwell, 2000), show that the commonest species were 

all found during this study. Only a few species larger 

than 5 mm were still to be found since the asymptotes 

are reached early in the accumulation. In fact, 30 samples, 

rather than the 65 taken, would have probably sufficed 

to record the same number of species. When the four 

small species are included the asymptote is not reached. 

Obviously, some species smaller than 5 mm were 

missed. 

The species accumulation curve computed with the 

25 samples and 53 species of the 1995 study shows that 

this inventory comes close to finding all the species in the 

area. However, the curve does not reach a clear asymptote 

(fig. 2) and some rare species are still missing. The 

incorporation of the five sub-fossil species into the list, 

by inflating the number of rare species, strongly increases 

the slope of the curve. 

Figure 2. Accumulation curves computed after 500 iterations: 1989a, 

all species collected during the 1989 study; 1989b, as 1989a but excluding four 

species smaller than 5 mm; 1996a, all species collected during the 1996 study; 

1996b, as 1996a but excluding the five hygrophilous species. 


Species CODES 1989 

1995 

Total Visual 

Acanthinula aculeata (Müller, 1774) AAC (1) 3 894 

Abida polyodon (Draparnaud, 1801) APO 11 48 

Bythiospeum sp. BYT (2) 

Cecilioides (Cecilioides) acicula (Müller, 1774) CAC 2 891 

Chondrina avenacea (Bruguière, 1792) CAV 1 

Ciliella ciliata (Hartmann, 1821) CCI 12 

Candidula gigaxii (Pfeiffer, 1850) CGI 334 264 13 

Cepaea (Cepaea) nemoralis (Linnaeus, 1758) CNE 4 15 9 

Cochlicella acuta (Müller, 1774) COA 213 

Cochlicella barbara (Linnaeus, 1758) COB 7 

Cochlostoma (Turritus) patulum (Draparnaud, 1801) CPA 1 329 

Carychium tridentatum (Risso, 1826) CTR (3) 

Candidula unifasciata (Poiret, 1801) CUN 57 295 

Cernuella (Microxeromagna) vestita (Rambur, 1868) CVE 95 9 

Cernuella (Cernuella) virgata (da Costa, 1778) CVI 48 1 051 32 

Euconulus (Euconulus) fulvus (Müller, 1774) EFU 846 

Ena (Ena) obscura (Müller, 1774) EOB 20 310 

Eobania vermiculata (Müller, 1774) EVE 90 302 156 

Granopupa granum (Draparnaud, 1801) GGR (6) 2 615 

Galba truncatulla (Müller, 1774) GTR (683) 

Granaria variabilis (Draparnaud, 1801) GVA 37 624 

Helix (Cornu) aspersa (Müller, 1774) HAS 103 114 76 

H. conspurcata or C. vestita HCV 

Hygromia cinctella (Draparnaud, 1801) HCI 6 1 

Hellicella (Xerotricha) conspurcata (Draparnaud, 1801) HCO 13 74 

Helicigona lapicida (Linnaeus, 1758) HLA 1 1 

Helix (Helix) melanostoma (Draparnaud, 1801) HME 34 6 6 

Jaminia (Jaminia) quadridens (Müller, 1774) JQU 32 118 

Lauria (Lauria) cylindracea (da Costa, 1778) LCY 339 

Monacha (Monacha) cantiana (Montagu, 1803) MAN 65 199 18 

Monacha (Monacha) cartusiana (Müller, 1774) MAR 8 246 20 

Oxychilus (Oxychilus) draparnaudi (Beck, 1837) ODR 64 513 8 

Oxychilus (Oxychilus) hydatinus (Rossmässler, 1838) OHY 20 

Pomatias elegans (Müller, 1774) PEL 643 1 997 883 

Phenacolimax (Phenacolimax) major (Férussac, 1807) PMA 182 

Pupilla (Pupilla) muscorum (Linnaeus, 1758) PMU 37 

Punctum (Punctum) pygmaeum (Draparnaud, 1801) PPY 1 434 

Pyramidula rupestris (Draparnaud, 1801) PRU 11 

Papilifera solida (Draparnaud, 1805) PSO 97 364 8 

Pseudotachea splendida (Draparnaud, 1801) PSP 12 129 129 

Sphincterochila (Albea) candidissima (Draparnaud, 1801) SCA 33 26 

Succinea (Succinella) oblonga (Draparnaud, 1801) SOB (1) 

Solatopupa similis (Bruguière, 1792) SSI 56 228 9 

Truncatellina callicratis (Scacchi, 1833) TCA (1) 2 246 

Trochoidea (Trochoidea) elegans (Gmelin, 1791) TEL 188 1 066 21 

Testacella (Testacella) haliotidea (Draparnaud, 1801) THA 2 

Trichia (Trichia) hispida (Linnaeus, 1758) THI 16 

Theba pisana (Müller, 1774) TPI 29 986 456 

Trochoidea (Trochoidea) pyramidata (Draparnaud, 1805) TPY 3 

Trochoidea (Trochoidea) trochoides (Poiret, 1789) TTR 43 1 586 106 

Perforatella (Monachoides) ventouxiana (Forcart, 1946) UVE 118 5 

Vallonia costata (Müller, 1774) VAC 2 806 

Vitrea (Crystallus) contracta (Westerlund, 1871) VCO (1) 3 807 

Vallonia enniensis (Gredler, 1856) VEN (75) 

Vallonia pulchella (Müller, 1774) VPU 438 

Vertigo (Vertigo) pygmaea (Draparnaud, 1801) VPY 811 

Cernuella (Xeromagna) cespitum (Draparnaud, 1801) XCE 94 153 115 

Xeropicta derbentina (Krynicki, 1836) XDE 600 7 017 639 

Zonites (Zonites) algirus (Linnaeus, 1758) ZAL 102 60 46 

Total number of shells 3 007 41543 2 778 

Table 1. Names, codes and number 

of individuals of the 58 species collected 

during the two studies. 

Numbers for species that were excluded 

from statistical analyses are given 

in parentheses. Nomenclature follows 

Kerney et al. (1983). 

The distinction between individuals 

collected during the visual search 

and the total is made for the 1995 study. 

69 



70 

Ecological results 

1989 study 

The four small species were excluded from the 

correspondence analysis, which was therefore applied 

to 28 species and 65 samples. The first two axes explain 

34.2 % of the total variation (λ = 2.21414) (fig. 3a). 

On axis 1 the species with negative values include 

Xeropicta derbentina, Trochoidea elegans and Theba 

pisana, and those with positive values include Pomatias 

elegans, Oxychilus draparnaudi, Cepaea nemoralis and 

Ena obscura. The first group of species is characteristic 

of dry and open habitats with herbaceous vegetation 

or ruderal environments. The second group of species 

is characteristic of open woodlands or environments 

with abundant leaf-litter. On axis 2 the species with the 

most positive values include Pomatias elegans, Oxychilus 

draparnaudi, Cepaea nemoralis, Monacha cantiana and 

Zonites algirus, and those with the most negative values 

include Candidula unifasciata, Papillifera solida, Granaria 

variabilis and Jaminia quadridens. 

On the same factorial plane, samples from cultivated 

and herbaceous wastelands are distinguished on axis 1 

(negative values) from woodland samples (positive 

values). Axis 2 separates samples taken in garrigues and 

grassland with chamaephytes (negative values) from 

samples taken from woodland. 

This analysis shows succession patterns of land snail 

communities in relation to the succession patterns of 

plant communities, from cultivation to woodland. For 

vegetation, two dynamic trajectories exist depending on 

perturbations or degradation of soil. Without perturbation 

and on thick soil, the main trajectory links cultivated areas 

to broad-leaved woodlands, through grassy fallow lands 

and closed scrublands. This trajectory corresponds to the 

upper part of the first factorial plane. If perturbations 

(such as fire) occur, another trajectory leads to garrigues 

through a continuous degradation of the terraces and 

soils. This trajectory corresponds to the lower part of the 

first factorial plane. Land snail communities, under the 

influence of vegetation structure, are strongly dependent 

on the age of abandonment of the terrace. They follow 

the two trajectories taken by plant communities along 

secondary successions. 

1995 study 

The five hygrophilous species were excluded from the 

correspondence analysis, which was therefore applied to 

53 species and 25 samples. The first three axes explain 

56.6 % of the total variation (λ = 1.32682) with 33.7, 12.3 

and 10.6 % respectively. 

A strong arch effect is apparent on the first factorial 

plane of the correspondence analysis (not shown here). 

Axes 1 and 2 represent the same strong contrast as in the 

analysis of the 1989 data between two very different types 

of malacological communities. They contrast open-land 

species, such as Cochlicella acuta, C. barbara, Trochoidea 

elegans, Theba pisana and Trochoidea trochoides to species 

characteristic of woodland, such as Acanthinula aculeata, 

Euconulus fulvus and Cepaea nemoralis. This arch effect 

results from the presence of two samples collected in 

agricultural crops. Despite the removal of the five subfossil 

species from these agricultural crop samples, their 

modern fauna is still very different from that of other 

open-land samples. Indeed, the irrigation of these crops 

apparently leads to a fauna characteristic of humid open 

lands of southeast France such as Cochlicella spp., Trichia 

hispida, Vallonia pulchella and Vertigo pygmaea. 

The second factorial plane of this analysis (axes 1 

and 3; fig. 3b) is similar to the first factorial plane of 

the 1989 study (fig. 3b). Secondary successions are the 

main structuring factor on this plane. Axis 3 distinguishes 

saxicolous species found in garrigues, such as Jaminia 

quadridens, Pseudotachea splendida, Helicigona lapicida and 

Chondrina avenacea, with negative values, from litterdwelling 

and shade-loving species found in woodland, 

such as Punctum pygmaeum, Euconulus fulvus and Ciliella 

ciliata, with positive values. Intermediate stages of the 

succession are found in the centre of the plane, where 

open land species, such as Cernuella cespitum, Candidula 

gigaxii and Granopupa granum, or open woodland species, 

such as Pomatias elegans, are found. 

DISCUSSION 

Comparison of the two sampling methods 

In studies of terrestrial gastropods, the area needed for 

adequate sampling and the number of samples required 

from each plot has long been a source of debate (e.g. 

Bishop, 1977; Menez, 2001). In his review stressing 

the need for quantitative sampling of molluscan faunas, 

Bishop (1977) suggested that the size of a quadrat should 

be less than 1 000 m² and that 30 (25 x 25 cm) units 

should be sampled within a 10 x 10 m quadrat. Menez 

(2001) considered that, in order to recover the entire 

molluscan fauna within a 1 km 2 square of Mediterranean 



71 

Figure 3. Correspondence 

analysis applied to the species/sample data matrix, open 

circles are samples and closed circles are species a) of the 

1989 study, axes 1 & 2; b) of the 1995 study, axes 1 & 

3. Arrows are the two secondary succession trajectories 

leading to either ‘woodland’ or ‘garrigue’. 



72 

habitat, 2.5 hours of searching and 4.2 litres of soil and 

litter are necessary. However, such a sampling regime was 

not possible within the constraints of the present project, 

which aimed to identify variation in faunal composition 

over much larger areas rather than to obtain complete 

inventories. Several different methods can be employed 

depending on the aims of the study. For instance, Cameron 

(1986) sampled 5 litres of litter and soil within a 30 x 30 m 

square, whereas for another study (Cameron & Morgan- 

Huws, 1975), he sampled 10 small quadrats (20 x 20 cm, 

further reduced to 15 x 15 cm) within 100 m². In order 

to study small-scale variation in the molluscan fauna, 

Davies and Grimes (1999) sampled five 25 x 25 cm 

squares within an undefined area. In conclusion, the 

sampled area has to be large enough to find the majority 

of species present in a specific environment but not so 

large as to include too much heterogeneity, which would 

reduce the information provided. 

The sampling methods used in both studies were 

intensive and conservative considering the small areas 

involved. In 1989, 15 minutes searches were employed 

in quadrats of approximately 25 m 2 (middle of terrace). 

Using this method, few large species would have been 

missed within the quadrat. In 1995, 20 minutes searches 

were employed in 400 m 2 quadrats, and all snails in the 

litter and soil of a 0.3125 m 2 area were collected. 

It is impossible to know whether either method 

resulted in a complete inventory of the species that were 

being searched for (i.e. large species only in 1989 or 

all species in 1995). However, the difference between 

the species list generated in 1989 (25 m 2 searched for 

15 minutes) with that resulting from the visual search of 

1995 (400 m 2 searched for 20 minutes) indicates that 20 

minutes is not sufficient to record all the large species in 

a 400 m 2 area, whereas 15 minutes searching in 25 m 2 

resulted in missing few large species, when compared to 

the 1995 complete list of species. Therefore it seems that 

the enlargement of the sampling area in fact decreased 

sampling efficiency, and collection of litter and soil was 

necessary to recover the large species that were missed. 

The more rapid saturation of the accumulation curve 

of the 25 m 2 samples of the 1989 study is counterintuitive. 

Indeed less heterogeneity should be included 

within smaller areas and therefore more of these small 

areas should be sampled to obtain the same number of 

species. Two factors can explain this effect. First, the 1995 

study included more diverse habitat types, which led to 

a slower saturation of the curves. Second, for areas of 

this magnitude (between 25 and 400 m 2 ), micro-habitat 

heterogeneity is the most important factor in structuring 

land snail communities. Here, it seems that most of this 

heterogeneity is included in small 25 m 2 squares and that 

no more large species would be found in larger areas 

unless a completely different type of vegetation were 

present. This suggests that a quadrat size of 25 m 2 , similar 

to that adopted by Labaune and Magnin (2001, 2002), 

strikes a good balance between maximising the habitat 

heterogeneity sampled and minimising sampling effort. 

Quadrats of approximately this size are thus likely to be 

the most appropriate for studies of terrestrial molluscs. 

Methods to estimate diversity 

and shed light on factors structuring 

land snail communities 

The systematic sampling of the 1995 study has led 

to the description of various land snail communities 

occurring in different habitats, from irrigated agricultural 

cropland to closed woodland. The 1989 study of postcultural 

succession did not sample such extreme 

environments but all the intermediate stages of the 

succession were encountered. Each species maps to a 

similar place on the 1989 and 1995 factorial planes. 

Only Cepaea nemoralis maps to a different place on the 

two graphs. It falls among wooded habitats in the 1989 

study but in garrigues in the 1995 study. Indeed, this 

mostly litter-dwelling species is known to occur in various 

environments (Kerney et al., 1983). 

The two methods revealed the importance of the 

structure of vegetation as well as that of the nature of 

ground cover. However, collection of soil samples at 

regular intervals allowed the recovery of many more 

species and a more accurate inventory of species 

diversity within the parish. Both studies showed that it 

is not necessary to collect an extremely large number of 

samples to have a good image of the factors influencing 

species distribution at this scale. However, the exploration 

of rare and spatially restricted habitat such as riparian 

forests seems necessary if the complete list of species of 

such an area is to be known. Neither of the present studies 

included such habitats and the total of 58 species is an 

underestimate of the total richness of the area. 

The sampling strategies differed between the two 

studies not only in the size of the quadrats but also in 

the initial aims and protocols. In 1989, only a few slopes 

in the northern and western part of the area were sampled 

but many samples were collected, whereas in 1995, the 

whole parish was covered but only a few samples were 

collected. The 1989 study aimed to describe molluscan 

communities along secondary successions and in order 



to study this structuring factor, sampling controlled for 

the stage of succession where land snails were collected. 

The 1995 study aimed to explore molluscan diversity at 

different spatial scales within the parish. An important 

result is that both studies revealed the same pattern of 

species distribution and the same powerful structuring 

factor, that is, secondary successions. This factor applies 

as much to the entire community as to a section of 

the community (i.e. large species) and highlights the 

strong influence of humans on the present landscape. 

Furthermore, the geographical position of the samples 

from the two studies sheds light on the landscape 

organisation of the studied area. At present, agricultural 

crops are found at lower altitude, close to the village 

centre and the river, whereas habitats at the end of the 

succession are found at higher altitude and further from 

the village. 

The ‘parish’ or ‘finage’ is a critical 

scale for landscape ecology 

Landscapes are forged by human activities (Forman 

& Godron, 1986). In the context of a rural French 

landscape, the scale of the parish is the most appropriate 

at which to describe this landscape and shed light 

on factors structuring it. The ‘parish’ (or ‘finage’ in 

French) is a well-delimited territory managed by a 

human community or village. This territory is divided 

into zones of different land uses, creating a mosaic of 

different elements (habitations, agricultural crops, pasture 

and forest). These elements are organised according to 

the topography, the agricultural potential of the area and 

the distance from the village. Thus, habitations are usually 

found close to a river or to any water source and are 

most closely surrounded agricultural crops, then pastures, 

and most distantly, forests. For a given region, the same 

pattern of land-use is found in all parishes, creating a 

mosaic of these repeated elements. 

This structure has existed for centuries. In the 

Mediterranean region, during Antiquity, three zones were 

defined: ‘sylva’ (managed forest) ‘ager’ (agriculture) and 

‘saltus’ (scrublands and pasture). Because of the difficulty 

of access and the steep slopes that precluded agriculture, 

‘sylva’ corresponded to zones at higher altitude and 

distant from habitations. ‘Ager’ surrounded habitations 

and ‘saltus’ was where the soil could not support crops. 

The modern landscape of the Auriol parish is similar in 

structure, although this similarity does not result directly 

from the antique structure, as it has undergone several 

changes. Under the pressure of human population 

increase, agriculture covered the entire area in the 

nineteenth century. Later, following the mechanization 

of agriculture and the movement of population to towns 

most of this land was abandoned (Grove & Rackham, 

2001) and left to ecological succession. Furthermore, 

the modern landscape, although similar to that of the 

Antiquity in appearance, supports different functions. 

Today’s landscape structure at Auriol is explained in 

terms of secondary succession following abandonment 

of terraces because of the greater difficulty of accessing 

them. Garrigues correspond to poor stony soils and more 

arid environments and can be created by perturbations 

such as fire. In terms of secondary succession, these 

garrigues are usually found on degraded terraces with 

eroded soil. However, they also result from human 

activities. Historically, they result from the opening of 

the environment by cattle and the collection of woody 

plants. 

The activities of the village, as a human community, 

structure the landscape. They are organised according 

to the interaction between soil depth, access to water 

sources and topography. This explains why both studies 

show the same pattern. By studying abandoned terraces 

along a few slopes, the 1989 study in fact studied the 

main structuring factor of the ‘finage’, while the 1995 

study revealed the spatial structure of this factor at the 

‘finage’ scale. 

Each zone defined by human activities offers 

peculiar vegetation structure and microhabitats to which 

molluscan communities are responding closely. Land snail 

communities are the result of historic management and 

are influenced by the recent evolution of the landscape, 

which reconstructs, to some extent, the antique spatial 

model of land use. This recent evolution is taking two 

paths: reforestation following land abandonment and 

urbanisation. The 1995 study highlighted the importance 

of the former path in structuring land snail communities 

at the scale of the parish. However, the importance of the 

latter will certainly grow dramatically in the future. 


The authors thank Errol Vela, Sébastien Della Casa 

and Bernard Hill for their help in the field. They also 

thank Dr R.H. Cowie, Dr N. Pike and an anonymous 

referee for their helpful comments on an earlier version 

of the manuscript. This work was supported by the 

programme of the EGPN (Écologie et gestion du 

73 



74 

patrimoine naturel) committee of the French Ministry 

of Environment « Dynamique de la biodiversité et 

action de l’homme » programme: « Changement dans la 

structuration du paysage et dynamique de la biodiversité 

en région méditerranéenne – Approche à différentes 

échelles d’espace et de temps ». 

References 

BISHOP M.J., 1977. Approaches to the quantitative description 

of terrestrial mollusc populations and habitats. Malacologia, 

16: 61-66. 

BOAG D.A., 1982. Overcoming sampling bias in studies of terrestrial 

gastropods. Canadian Journal of Zoology, 60: 1289- 

1292. 

BOYCOTT A.E., 1934. The habitats of land Mollusca in Britain. 

Journal of Ecology, 22: 1-38. 

CNRS, 1975. Carte climatique détaillée de la France. Feuille de 

Marseille. Ophrys, Gap. 

CAMERON R.A.D., 1986. Environment and diversities of forest 

snail faunas from coastal British Columbia. Malacologia, 27: 

341-355. 

CAMERON R.A.D., 2002. The land molluscs of North Ronaldsay, 

Orkney: human intervention and island faunal diversity. 

Journal of Conchology, 37: 445-453. 

CAMERON R.A.D. & MORGAN-HUWS D.I., 1975. Snail faunas 

in the early stages of a chalk grassland succession. Biological 

Journal of the Linnean Society, 7: 215-229. 

COLWELL R.K., 2000. Estimates: Statistical estimation of species 

richness and shared species from samples (Software and user’s 

guide) In: http://viceroy.eeb.uconn.edu/estimates 

DAVIES P. & G RIMES C.J., 1999. Small-scale spatial variation 

of pasture molluscan faunas within a relic watermeadow 

system at Wylye, Wiltshire, U.K. Journal of Biogeography, 26: 

1057-1063. 

EMBERTON K.C., PEARCE T.A. & RANDALANA R., 1996. 

Quantitatively sampling land snail species richness 

Madagascan rainforests. Malacologia, 38: 203-212. 

FORMAN R.T.T. & GODRON M., 1986. Landscape Ecology. John 

Wiley & Sons, New York, 619 pp. 

GROVE A.T. & RACKHAM O., 2001. The nature of Mediterranean 

Europe, an ecological history. Yale University Press, New 

Haven, 384 pp. 

HAWKINS J.W., LANKESTER M.W. & NELSON R.R.A., 1998. 

Sampling terrestrial gastropods using cardboard sheets. 

Malacologia, 39: 1-9. 

KERNEY M.P., CAMERON R.A.D. & JUNGBLUTH J.H., 1983. Die 

landschnecken Nord und Mitteleuropas: fremdländische gewächschausarten. 

Verlag, Berlin, 384 pp. 

LABAUNE C. & MAGNIN F., 2001. Land snail communities in 

Mediterranean upland grasslands: the relative importance of 

four sets of environmental and spatial variables. Journal of 

Molluscan Studies, 67: 463-474. 

LABAUNE C. & MAGNIN F., 2002. Pastoral management vs. 

land abandonment in Mediterranean uplands: impact on 

land snail communities. Global Ecology and Biogeography, 

11: 237-245. 

LEGENDRE L. & LEGENDRE P., 1984. Ecologie numérique. Masson 

& Presses de l’Université du Québec, Paris & Montréal. 

MAGNIN F., 1991. Mollusques continentaux et histoire quaternaire 

des milieux méditerrannéens (Sud-Est de la France, 

Catalogne) Ph.D. Thesis, UFR des Sciences géographiques 

et de l’aménagement, Aix-Marseille II, 340 pp. 

MAGNIN F. & TATONI T., 1995. Secondary successions on 

abandoned cultivation terraces in calcareous Provence. II. 

The gastropod communities. Acta Œcologica, 16: 89-101. 

MAGNIN F., TATONI T., ROCHE P. & BAUDRY J., 1995. Gastropod 

communities, vegetation dynamics and landscape changes 

along an old-field succession in Provence, France. Landscape 

and Urban Planning, 31: 249-257. 

MENEZ A., 2001. Assessment of land snail sampling efficacy 

in three Mediterranean habitat types. Journal of Conchology, 

37: 171-175. 

TATONI T., MAGNIN F., BONIN G. & VAUDOUR J., 1994. 

Secondary successions on abandoned cultivation terraces in 

calcareous Provence. I. Vegetation and soil. Acta Œcologia, 

15: 431-447. 

TER BRAAK C.F.J. & SMILAUER P., 1998. CANOCO 4.0. GLW- 

CPRO, Wageningen, The Netherlands. 


Évolution de la croissance radiale du pin d’Alep (Pinus 

halepensis Mill.) en Provence calcaire (sud-est de la France) 

Evolution of Aleppo pine (Pinus halepensis Mill.) 

radial growth in calcareous Provence (South-Eastern France) 

Cyrille B. K. Rathgeber 1 , Antoine Nicault 2 & Joël Guiot 3 

1. Auteur pour correspondance : Centre INRA de Nancy, UMR- INRA-ENGREF 1092, Laboratoire d’étude de la ressource forêt-Bois (LERFoB), 

Équipe qualité des bois, 54280 Champenoux, France. Tél. : +33 3 83 39 40 64 ; Fax : +33 3 83 39 40 69. Courrier électronique : cyrille.rathgeber@ 

nancy.inra.fr 

2. Centre d’études nordiques (CEN), Université Laval, Sainte-Foy, Québec, G1K 7P4, Canada 

3. CEREGE, CNRS-UMR 6635, BP80, Europôle méditerranéen de l’Arbois, 13545 Aix-en-Provence Cedex 4, France 

Résumé 

Les changements planétaires (changements climatiques, augmentation 

du taux de CO 2 atmosphérique et augmentation des 

dépôts azotés) sont susceptibles d’avoir un effet sur la production 

des écosystèmes forestiers. La dendrochronologie offre un moyen de 

vérifier cette hypothèse en milieu naturel, car les données de croissance 

radiale peuvent être interprétées comme des indicateurs de la 

productivité des écosystèmes forestiers. Les données dendrochronologiques 

utilisées dans cette étude proviennent de 21 peuplements 

de pin d’Alep localisés en Provence calcaire, dans le sud-est de la 

France. Les variables analysées sont : la densité minimale, la densité 

du bois initial, la largeur du bois initial, la densité maximale, la 

densité du bois final, la largeur du bois final, ainsi qu’un indice 

synthétique de croissance. Les tendances sont détectées sur la période 

1950-1990. Cette étude n’a pas permis de mettre en évidence de 

tendances générales positives et significatives concernant les variables 

relatives à la largeur des cernes, contrairement à ce que laissait 

prévoir la littérature scientifique. Ce travail a cependant permis de 

mettre en évidence une augmentation de la densité moyenne du bois 

initial et une diminution de la densité moyenne du bois final ainsi 

que de la densité maximale. Ces résultats montrent que la baisse de 

densité du bois final observée par différents auteurs pour le nord 

et le centre de l’Europe semble également être valide pour le sud de 

l’Europe. Cela suggère l’influence d’un facteur commun à toutes ces 

régions qui pourrait être l’augmentation des dépôts azotés. 

Mots-clés 

Changements planétaires, dendrochronologie, densité du bois, 

pin d’Alep, Provence 

Abstract 

Global changes: climatic change, atmospheric CO 2 rise and nitrogen 

deposition increase, are likely to have an effect on forest production. 

Dendrochronology offers the possibility to verify this hypothesis in 

natural conditions since tree-ring data can be interpreted as indicators 

of forest ecosystem productivity. Numerous studies of radial 

tree growth have been carried out, mainly in temperate and cold 

climate regions. However, the Mediterranean region appears particularly 

promising for researching the causes of the observed trends. 

Dendrochronological data used in this study come from 21 Aleppo 

pine stands located in South-Eastern France. Analysed variables 

are: minimal density, earlywood density, earlywood width, maximal 

density, latewood density, latewood width, and a tree growth index. 

Raw data are used for minimal and maximal density, whilst the 

other variables are standardised. Trends are detected using simple 

linear regressions over the period from 1950 to 1990. In contrast to 

the results published in the scientific literature, this study does not 

find out positive general significant trends regarding ring widths. 

However, this work shows an increase of earlywood density and a 

decrease of latewood and maximum density. These observations 

are coherent with some recent experimental results, suggesting that 

the increase of the earlywood density could be a consequence of 

CO 2 increase, whilst the decrease of latewood density could be a 

consequence of the increase of nitrogen deposition. Moreover, these 

results show that the decrease of latewood density, observed by several 

authors for North and Central Europe, seems also to be pertinent for 

South Europe. This suggests the influence of a common factor to all 

these regions which could be the increase of nitrogen deposition. 

Key-words 

Global change, tree-ring, wood density, Aleppo pine, 

French Mediterranean region 

75 


◆ C.B.K. RATHGEBER, A. NICAULT & J. GUIOT 

76 

INTRODUCTION 

L’augmentation du taux de CO 2 atmosphérique, depuis 

une concentration préindustrielle de 280 µmol.mol -1 

jusqu’à une concentration actuelle de 370 µmol.mol -1 

(Keeling & Worf, 2001), est susceptible d’avoir un effet 

fertilisant sur les écosystèmes forestiers (Hättenschwiler 

et al., 1997 ; Curtis, 1998 ; Norby et al., 1999). Son 

action peut être directe : l’augmentation du substrat le 

plus limitant de la photosynthèse devrait permettre une 

accélération de cette dernière ; et indirecte : limitation de 

la photorespiration (Mousseau, 1998) et amélioration de 

l’efficience de l’utilisation de l’eau (Scarascia-Mugnozza 

& Angelis, 1998). De plus, les changements climatiques, 

provoqués par l’augmentation des gaz à effet de serre, 

ainsi que la pollution atmosphérique sont également 

susceptibles de modifier la productivité des écosystèmes 

forestiers. 

La largeur et la densité des cernes annuels de croissance 

peuvent être considérées comme des indicateurs de 

la productivité des écosystèmes forestiers (Graumlich et 

al., 1989). Ainsi, de nombreuses analyses de la tendance 

à long terme ont été réalisées sur des séries de largeurs 

de cernes provenant de peuplements de feuillus ou de 

résineux, principalement en France, Suisse, Allemagne 

et Amérique du Nord (Innes, 1991 ; Graybill & Idso, 

1993 ; Badeau et al., 1996 ; Spiecker et al., 1996). La 

détection et l’analyse des tendances à long terme en 

région méditerranéenne apparaissent particulièrement 

intéressantes dans la recherche actuelle de l’attribution 

des causes responsables des augmentations de croissance 

observées et de leurs conséquences pour l’avenir. En effet, 

en région méditerranéenne les arbres sont généralement 

plus sensibles à l’effet du bilan hydrique qu’à celui du 

bilan thermique (Cherubini et al., 2003). Les ligneux 

devraient donc également être plus sensibles à l’effet 

de l’augmentation du taux de CO 2 (amélioration de 

l’efficience de l’utilisation de l’eau) qu’à l’effet de l’augmentation 

des températures (Feng et al., 1999 ; Knapp et 

al., 2001). Malgré cela, peu de travaux ont été consacrés 

à cette région (Scarascia-Mugnozza et al., 2000). Il est 

également intéressant de noter que peu d’études sur la 

tendance à long terme prennent en compte la densité du 

bois. Or, l’utilisation des différentes variables de densité 

devrait permettre de mieux apprécier la répartition des 

augmentations de productivité au cours de l’année et ainsi 

de mieux analyser leurs causes. 

Cet article a donc pour objectif : (1) de détecter 

l’existence potentielle de tendances à long terme dans 

les séries de largeur et de densité de cernes provenant de 

21 peuplements de pins d’Alep (Pinus halepensis Mill.) 

situés en Provence calcaire (sud-est de la France) ; (2) 

d’analyser les causes possibles des tendances mises en 

évidence ; et (3) de confronter les résultats obtenus à la 

fois aux résultats d’autres études dendrochronologiques 

et aux résultats d’expériences réalisées sous conditions 

contrôlées en laboratoire. 

MATÉRIEL ET MÉTHODES 

Les données dendrochronologiques utilisées dans cette 

étude proviennent de 21 peuplements de pin d’Alep localisés 

en Provence calcaire, dans le sud-est de la France 

(fig. 1). Ces données sont présentées en détail dans 

les thèses de Nicault (1999) et Rathgeber (2002). Les 

variables analysées au cours de la présente étude sont : la 

densité minimale (ND), la densité du bois initial (ED), la 

largeur du bois initial (EW), la densité maximale (XD), la 

densité du bois final (LD), la largeur du bois final (LW) ; 

ainsi qu’un indice de croissance synthétique (TGI), défini 

comme suit : TGI = EW×ED + LW×LD. (1) 

L’intérêt de ce dernier indice est de mieux prendre en 

compte la production de biomasse en combinant à la fois 

les variables de largeur et de densité (Hättenschwiler & 

Körner, 1996 ; Rathgeber et al., 2000). 

Dans cette étude, les variables ND et XD n’ont pas 

été standardisées car elle sont moins sensibles à l’effet 

de l’âge que les autres variables dendrochronologiques 

(Schweingruber, 1990). Les variables ED, EW, LD et 

LW sont tout d’abord standardisées par un indice de 

productivité spécifique à chaque site ce qui permet d’éliminer 

les différences de fertilité entre les peuplements 

(voir Rathgeber et al., 2000). L’indice utilisé est calculé en 

faisant la somme des 50 premières valeurs de la variable à 

standardiser. Cette méthode est communément employée 

pour standardiser les largeurs de cernes à l’aide du rayon 

de l’arbre à 50 ans (Rathgeber et al., 1999). Ces variables 

sont ensuite standardisées à l’aide de la loi biologique 

afin d’éliminer l’effet de l’âge (Becker, 1989). Le TGI 

est calculé à partir des variables standardisées. Enfin, 

une chronologie régionale valable pour l’ensemble de 

l’aire d’étude est calculée pour chacune des variables 

dendrochronologiques en moyennant les chronologies 

des 21 populations. 

La tendance à long terme est examinée sur la période 

1950-1990, pour chaque variable et pour chaque population, 

à l’aide de régressions linéaires simples en fonction 

du temps. Cette période est cependant marquée par un 


ÉVOLUTION DE LA CROISSANCE RADIALE DU PIN D’ALEP EN PROVENCE ◆ 

Figure 1. Localisation des 21 peuplements de pin d’Alep. 

Les points indiquent les peuplements et le carré noir la station 

météorologique. Le code des peuplements fait référence au tableau 1. 

Figure 1. Locations of the 21 Aleppo pine stands. 

Black dots mark the stands and the square represents 

the meteorological station. Stands code refers to table 1. 

77 

phénomène climatique d’une ampleur exceptionnelle : 

le gel de février 1956. Cet événement a très fortement 

réduit la croissance de l’année en cours et même parfois 

des années suivantes (Devaux & Le Bourhis, 1978). Nous 

avons préféré ne pas le prendre en compte dans l’analyse 

des tendances et retirer systématiquement les années susceptibles 

d’être affectées (1956, 1957, 1958 et 1959). 

Les données météorologiques utilisées dans ce travail 

ont été fournit par Météo France et proviennent de l’observatoire 

du Palais Longchamp à Marseille. À partir de 

ces données, les tendances climatiques observées au cours 

du XX e siècle en Provence ont été décrites par Rathgeber 

(2002). 

Les relations cernes-climat concernant une partie des 

variables utilisées ont été analysées par Nicault (1999) et 

Rathgeber et al. (2000) à l’aide de fonctions de réponse 

(Guiot, 1991). À partir de ces travaux, nous déterminerons 

si les changements climatiques observés peuvent être 

rendus responsables des tendances mises en évidence sur 

la croissance. 



RÉSULTATS 

À l’échelle régionale, on observe pour le bois initial une 

diminution de la densité minimale (fig. 2a) alors que la 

densité moyenne ne présente pas de tendance significative 

(tableau 1). Pour le bois final, on observe une diminution 

très nette de la densité minimale et de la densité moyenne 

(fig. 2b). La série régionale de TGI ainsi que les largeurs 

de bois initial et final ne présentent pas de tendance significative 

lorsque l’on considère l’ensemble de la période 

1950-1999. Mais cette période peut être décomposée en 

deux phases distinctes : (1) une période de relative stabilité 

allant de 1950 à 1976 et (2) une période de forte 

diminution allant de 1977 à 1990 (fig. 2c). 

Au niveau local, on observe une tendance générale à 

l’augmentation pour la densité du bois initial : 9 populations 

présentent des tendances significatives positives 

alors qu’une seule population présente une tendance 

significative négative (tableau 1). La densité minimale, 

en revanche, présente des tendances négatives significatives 

pour 7 populations sans présenter aucune tendance 

positive. En ce qui concerne la largeur du bois initial, 7 

populations montrent des tendances significatives négatives 

alors que 2 populations seulement montrent des 

tendances significatives positives. Une forte tendance 

générale à la baisse se manifeste pour les variables de 

densité du bois final. En ce qui concerne la densité du 

bois final, 12 populations présentent des tendances néga- 

78 

Peuplements 

ND EDS EWS XD LDS LWS TGI 

α S a α S a α S a α S a α S a α S a α S a 

AUR NS *** + 2 NS *** - 1.36 ** - 0.61 NS NS 

CIO NS NS NS *** - 0.96 ** - 0.88 NS NS 

CVI ** - 0.19 NS NS *** - 1.53 * - 0.75 * - 9.0 NS 

FAR *** - 0.40 *** + 1 NS *** - 1.52 *** - 1.00 * + 14.0 * + 15.0 

FOO NS NS NS * - 0.43 NS NS NS 

GAR abs abs abs NS ** - 3.00 abs abs abs ** - 0.91 NS ** - 2.0 

MAL NS NS * - 2.00 *** - 1.62 NS NS ** - 4.0 

MAN *** - 0.64 *** - 1 NS *** - 1.54 *** - 0.00 NS ** - 11.0 

MEE NS ** + 0.78 NS *** - 0.81 NS NS NS 

MOU NS NS NS *** - 1.06 NS NS NS 

MUR NS NS NS *** - 1.23 *** - 0.60 *** - 9.0 ** - 13.0 

NYO NS *** + 2 ** - 6.00 *** - 1.42 NS * - 5.0 ** - 10.0 

PER *** - 0.93 NS *** + 18.00 ** - 0.57 NS *** + 16.0 *** + 33.0 

PNM *** - 0.67 NS NS *** - 1.46 ** - 1.00 NS NS 

RDA *** - 0.51 NS * + 4.00 *** - 1.21 NS NS * + 9.0 

RIA NS *** + 2 NS *** - 1.73 *** - 1.00 NS NS 

ROB *** - 0.38 NS *** - 7.00 *** - 2.13 *** - 2.00 *** - 13.0 *** - 20.0 

ROG NS *** + 2 ** - 7.00 *** - 1.65 NS NS NS 

ROU NS ** + 1 *** - 16.00 ** - 0.86 NS NS ** - 22.0 

SIM NS *** + 1 NS *** - 1.54 ** - 0.77 NS NS 

SIS NS ** + 0.56 *** - 10.00 *** - 2.96 *** - 3.00 ** - 12.0 *** - 24.0 

Régional ** - 0.18 NS 0.67 NS 0.17 *** - 1.32 *** - 0.88 NS 2.0 NS 2.00 

Total NS 13 11 12 0 9 14 10 

Total + 0 9 2 0 0 2 3 

Total - 8 1 7 21 12 5 8 

Tableau 1. Tendances à long terme observées. Pour chaque population et chaque variable dendrochronologique : densité minimale (ND), densité du bois initial (ED), largeur 

du bois initial (EW), densité maximale (XD), densité du bois final (LD), largeur du bois final (LW) et indice synthétique de croissance (TGI), la tendance est détectée sur la 

période 1950-1990 par régression linéaire simple. Les années 1956 à 1959 ne sont pas prises en compte. Le niveau de significativité α (NS pour p > 0.1, * pour p ≤ 0.1, 

** pour p ≤ 0.05 et *** pour p ≤ 0.01) de la tendance, ainsi que le signe s et la pente a (×10 3 ) de la droite de régression sont fournis en colonne. 

Table 1. Observed long term trend. For each population and each denrochronological variable : minimal density (ND), earlywood density (ED), earlywood width (EW), 

maximal density (XD), latewood density (LD), latewood width (LW) and tree growth index (TGI), trend is detected on the 1950–1990 period using simple linear regression. 

Years from 1956 to 1959 are not taken into account. Significance level α (NS for p > 0.1, * for p ≤ 0.1, ** for p ≤ 0.05 and *** for p ≤ 0.01) of the trend, as well 

as the sign s and the slope a (×10 3 ) of the regression line are given in column. 



79 

Fig. 2. Tendances régionales à long terme. 

a. Densité minimale (ND) et densité du bois initial (ED). 

b. Densité maximale (XD) et densité du bois final (LD). 

c. Largeur du bois initial (EW), largeur du bois final (LW) et indice synthétique 

de croissance (TGI) en relation avec le cumul des précipitations annuelles (P). 

Les statistiques relatives aux régressions linéaires (en pointillés) sont données 

dans le tableau 1. 

Fig. 2. Regional long term trends. 

a. Minimal density (ND) and earlywood density (ED). 

b. Maximal density (XD) and latewood density (LD). 

c. Earlywood width (EW), latewood width (LW) and tree growth 

index (TGI) in relation with total annual precipitation (P). 

Statistics of the regression lines (dotted) are presented in table 1. 



80 

tives significatives. Pour la densité maximale toutes les 

populations étudiées présentent des tendances négatives 

significatives. Pour la largeur du bois final, 5 populations 

présentent des tendances significatives négatives alors 

que 2 populations présentent des tendances significatives 

positives. En ce qui concerne le TGI, 3 populations 

présentent des tendances positives, alors que 8 populations 

présentent des tendances négatives. Les tendances 

positives sont globalement plus prononcées (pentes plus 

fortes) que les tendances négatives, elles sont pourtant 

moins significatives car elles sont souvent masquées par 

des variabilités interannuelles plus fortes. 

DISCUSSION 

Quatre facteurs principaux sont susceptibles d’être 

responsables des tendances observées : (1) l’effet de 

l’âge, (2) l’effet des changements climatiques, (3) l’effet 

de l’augmentation du taux de CO 2 et (4) l’effet de 

l’augmentation des dépôts azotés. Malgré la standardisation, 

l’effet de l’âge reste le premier facteur à prendre 

en compte pour expliquer les tendances à long terme 

observées sur les variables dendrochronologiques (Briffa 

et al., 1996). La méthode de standardisation utilisée dans 

cette étude donne de très bons résultats pour les largeurs 

(fig. 2a et 2c dans Rathgeber et al., 2000). L’ajustement 

de la loi biologique sous-estime en revanche légèrement 

la densité moyenne du bois initial pour les arbres de plus 

de 40 ans en raison de la forme de la courbe polynomiale 

retenue (fig. 2b dans Rathgeber et al., 2000). Cette sousestimation 

est susceptible d’induire une légère tendance 

positive dans les données standardisées. Cependant, en ne 

nous intéressant qu’à la période récente, nous avons limité 

ce problème qui ne devrait donc pas être responsable 

de la tendance à l’augmentation détectée pour la densité 

du bois initial. La loi biologique relative à la densité du 

bois final présente également des difficultés d’ajustement 

des données (fig. 2d dans Rathgeber et al., 2000). Ces 

difficultés se manifestent pour les classes d’âge de 50 à 

80 ans qui apparaissent comme hors norme et infléchissent 

fortement la courbe polynomiale. La loi biologique 

surestime donc la densité pour les âges compris entre 

50 et 80 ans, pour revenir à un ajustement de meilleure 

qualité ensuite. Ceci devrait donc induire une tendance 

positive dans les données standardisées et peut être rendu 

responsable du fait que l’on observe moins de tendances 

significatives négatives pour la densité du bois final que 

pour la densité maximale. 

Les changements climatiques ayant eu lieu au cours 

des 50 dernières années dans la région étudiée consistent 

en : (1) une augmentation générale des températures pour 

toutes les saisons, cette augmentation étant maximale en 

été ; et (2) une légère augmentation des précipitations 

hivernales alors que les précipitations des autres saisons 

sont restées stables (Rathgeber, 2002). Ces observations 

sont en accord avec une étude récente menée à l’échelle de 

la France (Lebourgeois et al., 2001). Une forte tendance 

à la baisse est cependant observable pour le cumul des 

précipitations annuelles entre 1977 et 1990 qui paraît être 

la cause de la baisse du TGI et de la largeur des bois initial 

et final sur cette même période. L’analyse des relations 

cernes-climat à l’aide des fonctions de réponses montre 

que la densité du bois initial et la densité minimale sont 

toutes deux influencées négativement par les précipitations 

des mois de mars et avril et les températures du mois 

de février (Nicault, 1999). Les précipitations relatives aux 

mois de mars et avril sont stables sur les cinquante dernières 

années alors que la température du mois de février 

a augmenté. L’influence du climat devrait donc induire 

une tendance négative pour ces variables. La tendance 

générale à la baisse observée pour la densité minimale 

peut donc être expliquée par l’évolution du climat alors 

que la tendance générale à la hausse observée sur la densité 

du bois initial ne peut pas l’être. La densité du bois 

final et la densité maximale sont toutes deux influencées 

positivement par les précipitations des mois d’avril et mai. 

Les précipitations printanières étant quasiment stables, 

le climat ne peut être tenu pour responsable des fortes 

tendances négatives observées pour ces variables. 

Deux études récentes analysent la réponse des variables 

dendrochronologiques à une augmentation du taux 

de CO 2 . La première étude (Hättenschwiler & Körner, 

1996) concerne de jeunes plants d’épicéa (Picea abies) 

soumis à un enrichissement en CO 2 et/ou en azote. Ces 

plants poussent dans une chambre fermée qui reproduit 

(du point de vue du sol, du climat, et de la compétition) 

les conditions de la station forestière dans laquelle ils 

ont été prélevés (Alpes suisses, 1 500 m d’altitude). La 

seconde étude concerne de jeunes plants de pins (Pinus 

taeda) soumis uniquement à un enrichissement en CO 2 et 

poussant dans des conditions nutritionnelles non limitantes 

(Telewski et al., 1999). En conditions nutritionnelles 

non limitantes on observe toujours un effet positif de 

l’enrichissement en CO 2 que ce soit sur la largeur ou bien 

sur la densité du cerne (ibidem). Or ce n’est pas le cas en 

conditions nutritionnelles limitantes, où l’effet du CO 2 

est le plus souvent moins important que celui de la fertilisation 

azotée (Hättenschwiler & Körner, 1996). Pour 



une augmentation simultanée de l’enrichissement en CO 2 

et en azote, Hättenschwiler et Körner (1996) décrivent 

(1) une très faible augmentation du TGI due à l’ensemble 

des effets du CO 2 et de l’azote sur les différentes 

variables ; (2) une très faible augmentation de la largeur 

totale due à une augmentation de la largeur du bois final ; 

(3) une augmentation de la densité du bois initial due 

à l’effet de l’augmentation du taux de CO 2 ; et (4) une 

diminution de la densité du bois final due à l’effet négatif 

de l’enrichissement en azote sur la densité. Les résultats 

obtenus par Hättenschwiler et Körner (1996) sont donc 

très proches des résultats obtenus par la présente étude, 

si l’on admet que de vieux arbres en conditions naturelles 

sont susceptibles de présenter des réponses plus faibles 

que de jeunes arbres en conditions expérimentales. 

La majeure partie des études dendrochronologiques 

sur la tendance à long terme présente des résultats mettant 

en évidence de fortes tendances positives qu’elles 

associent généralement à l’augmentation du taux de CO 2 

et à l’amélioration des conditions climatiques (Spiecker 

et al., 1996). De telles tendances n’ont pas été mises en 

évidence au cours de ce travail, alors même qu’il a été 

conduit en région méditerranéenne, région qui devrait 

être théoriquement particulièrement sensible à l’effet 

de l’augmentation du taux CO 2 . La baisse de la densité 

moyenne du bois final et de la densité maximale est cohérente 

avec les résultats de la littérature à l’échelle nationale 

(Bergès et al., 2000) comme à l’échelle mondiale (Briffa 

et al., 1998a ; Briffa et al., 1998b). L’étude précédente 

(Briffa et al., 1998b) concerne cependant essentiellement 

des peuplements de hautes latitudes et altitudes sensibles 

aux températures estivales ; alors que notre étude 

concerne des peuplements méditerranéens, de basses 

altitudes, et qui ne sont pas sensibles aux températures 

estivales mais aux précipitations printanières. Les raisons 

de cette baisse générale de la densité du bois final ne 

sont pas encore connues. Briffa et al. (1998b) proposent 

comme explication un changement dans la saisonnalité 

des variables climatiques. Vaganov et al. (1999) montrent 

que le retardement actuel de la fonte des neiges a une 

influence certaine en ce qui concerne les arbres de la forêt 

boréale. Notre étude quant à elle suggère qu’il existe une 

cause commune capable d’affecter des écosystèmes aussi 

différents que la forêt boréale et la forêt méditerranéenne 

et qui pourrait être l’augmentation des dépôts azotés. 

REMERCIEMENTS 

Ce travail a été financé par l’Union européenne, à travers 

le projet FORMAT (contrat ENV4-CT97-0641) et à 

travers une bourse individuelle post-doctorale Marie Curie 

(contrat EVK2-CT-2002-50021). Ce travail a également 

été soutenu par le ministère français de la Recherche à 

travers une bourse individuelle de thèse. Les données 

météorologiques ont été fournies par Météo France. 

Bibliographie 

BADEAU V., BECKER M., BERT D., DUPOUEY J.-L., LEBOURGEOIS F. 

& PICARD J.-F., 1996. Long-term growth trends of trees: 

ten years of dendrochronological studies in France. In: 

Spiecker H., Mielikaïnen K., Köhl M. & Skovsgaard J. 

(eds.), Growth trends in European Forest. Springer, Berlin, 

p. 167-181. 

BECKER M., 1989. The role of climate on present and past vitality 

of silver fir forests in the Vosges mountains of northeastern 

France. Canadian Journal of Forest Research, 19: 1110-1117. 

BERGÈS L., DUPOUEY J.-L. & FRANC A., 2000. Quercus petraea 

Liebl. in northern France since the middle of the nineteenth 

century. Trees-Structure and Function, 14: 398-408. 

BRIFFA K.R., JONES P.D., SCHWEIGRUBER F.H., KARLEN W. 

& SHIYATOV S.G., 1996. Tree-ring variables as proxyclimate 

indicators: problems with low-frequency signals. In: 

Jones P.D., Bradley R. S. & Jouzel J. (eds.), Climatic variations 

and forcing mechanism of the last 2000 years. Springer-Verlag, 

Berlin, Heidelberg, p. 9-41. 

BRIFFA K. R., SCHWEINGRUBER F. H., JONES P. D., OSBORN T. J., 

HARRIS I. C., SHIYATOV S. G., VAGANOV E. A. & GRUDD H., 

1998a. Trees tell of past climates: but are they speaking less 

clearly today? Philosophical Transactions of the Royal Society of 

London Series B-Biological Sciences, 353: 65-73. 

BRIFFA K.R., SCHWEINGRUBER F.H., JONES P.D., OSBORN T.J., 

SHIYATOV S.G. & VAGANOV E.A., 1998b. Reduced sensitivity 

of recent tree-growth to temperature at high northern latitudes. 

Nature, 391: 678-682. 

CHERUBINI P., G ARTNER B.L., TOGNETTI R., BRAKER O.U., 

SCHOCH W. & INNES J.L., 2003. Identification, measurement 

and interpretation of tree rings in woody species from 

Mediterranean climates. Biological Reviews, 78: 119-148. 

81 



82 

CURTIS P. S., 1998. A meta-analysis of leaf gas exchange and 

nitrogen in trees grown under elevated carbon dioxide. Plant 

Cell and Environment, 19: 127-137. 

DEVAUX J.-P. & LE BOURHIS M., 1978. La limite septentrionale 

du pin d’Alep en France. Étude dendroclimatologique de 

l’impact des froids exceptionnels. Revue de biologie et d’écologie 

méditerranéenne, 5 : 133-158. 

FENG X., CUI H., TANG K. & CONKEY L.E., 1999. D as an 

indicator of Asian Monsoon intensity. Quaternary Research, 

51: 262-266. 

GRAUMLICH L.J., BRUBAKER L.B. & GRIER C.C., 1989. Longterm 

trends in forest net primary productivity: Cascade 

Mountains, Washington. Ecology, 70: 405-410. 

GRAYBILL D.A. & IDSO S.B., 1993. Detecting the aerial fertilization 

effect of atmospheric CO 2 enrichment in tree-ring 

chronologies. Global Biogeochemical Cycles, 7: 81-95. 

GUIOT J., 1991. The bootstrapped response function. Tree-Ring 

Bulletin, 51: 39-41. 

HÄTTENSCHWILER S. & KÖRNER C., 1996. Effect of elevated 

CO 2 and increased nitrogen deposition on photosynthesis 

and growth of understorey plants in spruce model ecosystems. 

Œcologia, 106: 172-180. 

HÄTTENSCHWILER S., MIGLIETTA F., RASCHI A. & KÖRNER C., 

1997. Thirty years of in situ tree growth under elevated CO 2 : 

a model for future forest responses? Global Change Biology, 

3: 463-471. 

INNES J.L., 1991. High-altitude and high-latitude tree growth 

in relation to past, present and future global climate change. 

Holocene, 1: 168-173. 

KEELING C.D. & WORF T.P. (EDS.), 2001. Atmospheric CO 2 records 

from sites in the SIO air sampling network. Trends: a compendium 

of data on global change. Carbon Dioxide Information 

Analysis Center, Oak Ridge National Laboratory. Oak Ridge, 

Tennessee, USA. 

KNAPP P.A., SOULÉ P.T. & GRISSINO-MAYER H.D., 2001. Detecting 

potential regional effects of increased atmospheric 

CO 2 on growth rates of western juniper. Global Change 

Biology, 7: 903-917. 

LEBOURGEOIS F., GRANIER A. & BREDA N., 2001. An analysis of 

regional climate change in France between 1956 and 1997. 

Annals of Forest Science, 58: 733-754. 

MOUSSEAU M., 1998. Is dark respiration rate changed? In: 

Jarvis J. (ed.), European forest and global change. Cambridge 

University Press, Cambridge, p. 79-93. 

NICAULT A., 1999. Analyse de l’influence du climat sur les variations 

inter et intra-annuelles de la croissance radiale du pin 

d’Alep (Pinus halepensis Mill.) en Provence calcaire. Thèse 

de doctorat, université d’Aix-Marseille III, 254 p. 

NORBY R.J., WULLSCHLEGER S.D., GUNDERSON C.A., JOHNSON 

D.W. & CEULEMANS R., 1999. Tree response to rising CO 2 

in field experiments: implications for future forest. Plant Cell 

and Environment, 22: 683-714. 

RATHGEBER C., 2002. Impact des changements climatiques et 

de l’augmentation du taux de CO 2 atmosphérique sur la productivité 

des écosystèmes forestiers : exemple du pin d’Alep 

(Pinus halepensis Mill.) en Provence calcaire (France). Thèse 

de doctorat, université d’Aix-Marseille III, 276 p. 

RATHGEBER C., GUIOT J., ROCHE P. & TESSIER L., 1999. 

Augmentation de productivité du chêne pubescent en région 

méditerranéenne française. Annales des sciences forestières, 56 : 

211-219. 

RATHGEBER C., NICAULT A., GUIOT J., KELLER T., GUIBAL F. & 

ROCHE P., 2000. Simulated response of Pinus halepensis forest 

productivity to climatic change and CO 2 increase using a statistical 

model. Global and Planetary Change, 26: 405-421. 

SCARASCIA-MUGNOZZA G. & ANGELIS P.O.D., 1998. Is water 

use more efficiently? In: Jarvis P. G. (ed.), European Forest 

and Global Change. Cambridge University Press, Cambridge, 

p. 192-214. 

SCARASCIA-MUGNOZZA G., OSWALD H., PIUSSI P. & R ADO- 

GLOU K., 2000. Forests of the Mediterranean region: 

gaps in knowledge and research needs. Forest Ecology and 

Management, 132: 97-109. 

SCHWEINGRUBER F.H., 1990. Radiodensitometry. In: Cook E.R. 

& Kairiukstis L.A. (eds.), Methods of dendrochronology, applications 

in environmental sciences. Kluwer academic publisher, 

Dordrecht, p. 55-63. 

SPIECKER H., MIELIKA K., KÖHL M. & SKOVSGAARD J. (EDS.), 

1996. Growth trends in European forests. Studies from 12 

countries. European Forest Institute Research Report n° 5. 

Springer. Berlin. 

TELEWSKI F.W., SWANSON R.T., STRAIN B.R. & BURNS J.M., 

1999. Wood properties and ring width responses to long-term 

atmospheric CO 2 enrichment in field-grown loblolly pine 

(Pinus taeda L.). Plant Cell and Environment, 22: 213-219. 

VAGANOV E.A., HUGHES M.K., KIRDYANOV A.V., SCHWEIGRUBER 

F.H. & SILKIN P.P., 1999. Influence of snowfall and melt 

timing on tree growth in subartic Eurasia. Nature, 400: 

149-151. 


Dynamique de la végétation de mares temporaires 

en Afrique du Nord (Numidie orientale, NE Algérie) 

Vegetation dynamics in temporary ponds of North Africa 

(Oriental Numidia, NE Algeria) 

Gérard de Bélair 

Laboratoire de recherche sur les zones humides (LRZH), Université Badji Mokhtar, BP533, 23000 Annaba, Algérie. 

Email : gerard_de_belair@yahoo.com 

Résumé 

La dynamique écologique des milieux endoréiques temporaires 

méditerranéens a été peu étudiée en Afrique du Nord, et jamais en 

Numidie (NE algérien). Dans cette région, 26 mares temporaires 

ont fait l’objet de suivis phytoécologiques et hydrologiques et la 

végétation a été systématiquement échantillonnée chaque mois 

durant l’hydropériode, au cours de trois cycles annuels (1998- 

2001). Les stations étudiées sont distribuées le long de deux transects 

de 90 km environ. Une analyse inter-classes a été appliquée sur 

un tableau rassemblant 539 relevés et 136 espèces végétales. 

Cette analyse a pour but de séparer deux effets, l’un temporel 

qui permet de décrire la succession des espèces, l’autre spatial qui 

est relatif à la typologie stationnelle. Les résultats permettent de 

distinguer nettement deux périodes dans l’année (été/automne et 

hiver/printemps), définies notamment par la présence de Panicum 

repens d’une part et de Ranunculus baudotii d’autre part. La 

typologie aboutit à définir trois groupes parmi les mares, déterminés 

successivement par : groupe 1, Characeae et Juncus acutus ; 

groupe 2, Callitriche obtusangula et Isoetes velata subsp. typica ; 

groupe 3, C. obtusangula et Alisma plantago-aquatica. 

Une annexe comporte une description détaillée des 26 hydrosystèmes 

de mares temporaires et permet d’en percevoir la richesse floristique, 

incluant la présence remarquable d’un élément original représentatif 

d’une « poche relictuelle afrotropicale ». 

Mots-clés 

Mares temporaires, analyse inter-classes, succession végétale, 

typologie, biogéographie, élément afrotropical. 

Abstract 

The temporary ponds have been less studied in Northern Africa 

than in Mediterranean Europe. Moreover, no study is devoted to 

the ecology of these ponds in Numidia (North-Eastern Algeria). 

Thus to fill this gap, we have surveyed 26 temporary ponds and 

we have examinated the fluctuations of plant assemblages and 

hydrology. These samples were performed monthly (nine months) 

during the hydroperiod and along three annual cycles (1998- 

2001). The studied ponds are distributed along two transects of 

90 km. We obtain a table of 539 relevés and 136 plant taxa. 

This table was submitted to an ACP and then to a Discrimin : 

Between analysis. This analysis allows to distinguish temporal and 

spatial effects : the first describes the succession of taxa, and the 

second the typology of ponds. Our results show that the temporal 

effect conveyied two periods (summer-autumn and winter-spring) 

with respectively Panicum repens and Ranunculus baudotii. The 

spatial effect permits to divide the twenty-six temporary ponds into 

three groups: group 1, Characeae and Juncus acutus; group 2, 

Callitriche obtusangula and Isoetes velata subsp. typica; group 3, 

C. obtusangula and Alisma plantago-aquatica. 

An appendix includes a detailed description of these 26 temporary 

ponds, with some data related to plant richness and a focus about 

the most outstanding biogeographic elements present within the 

« afrotropical relict pocket ». 

Key-words 

Temporary ponds, discrimin: between analysis, plant succession, 

typology, biogeography, afrotropical element. 

83 


◆ G. DE BÉLAIR 

84 

INTRODUCTION 

La flore et la végétation des mares temporaires 

méditerranéennes ont fait l’objet de nombreux travaux 

et synthèses récents (Quézel, 1998 ; Médail et al., 

1998 ; Marti & Esteban, 2001 ; Rhazi et al., 2003 ; 

Grillas et al., 2004 ; Deil, 2005). Leur intérêt et leur 

originalité biogéographique et écologique ne sont plus 

à démontrer à l’échelle du bassin méditerranéen, mais 

les études phytodynamiques de ces milieux complexes 

demeurent peu nombreuses, particulièrement au sud de 

la Méditerranée alors que ces mares subissent des impacts 

anthropiques majeurs (Rhazi et al., 2001). 

L’Algérie, et plus particulièrement la Numidie 

(au sens des divisions biogéographiques proposées 

par Quézel & Santa, 1962-1963), est riche en mares 

temporaires répondant aux critères Ramsar (Grillas 

et al., 2004). Plus d’une centaine d’entre elles ont été 

échantillonnées de 1995 à 2001 (Samraoui & de Bélair, 

1997, 1998). Un certain nombre d’études avait été 

effectué auparavant sur l’Algérie de manière sporadique, 

incluant des travaux non seulement sur les Spermaphytes 

ou les Ptéridophytes, mais également sur les algues d’eaux 

douces : Cyanophycées, Diatomées, Myxophycées 

et Chlorophycées et, notamment, les Charophycées 

(Lefranc, 1865 ; Gauthier-Lièvre, 1937 ; Feldmann, 1946 ; 

Morgan, 1982 ; Stevenson et al., 1989). Gauthier-Lièvre 

(1937) présente dans sa thèse un historique très détaillé 

des travaux réalisés en Algérie et, plus particulièrement 

en Numidie, sur les zones humides. Cet auteur soulignait 

« qu’en Algérie, les conditions optimales pour le développement 

d’une flore riche et variée sont réalisées sur toute l’étendue du 

secteur numidien, car c’est là que la pluviosité est la plus élevée 

et que se trouvent sur de grandes surfaces des affleurements de 

terrains siliceux, en l’espèce des grès éocènes ». 

Tous ces motifs nous ont amené à choisir 26 mares 

temporaires de Numidie orientale, dispersées le long d’un 

axe de 90 km environ. Sur les cinq campagnes réalisées 

de 1996 à 2001, seules les trois dernières campagnes 

ont été retenues en raison de l’effort homogène 

d’échantillonnage exercé durant ces trois cycles (1998- 

2001) sur les 26 stations. L’objectif envisagé de cette 

étude était triple : (i) évaluer la biodiversité végétale de 

ces mares, (ii) étudier la succession des espèces dans le 

temps (structure temporelle), et (iii) définir la typologie 

stationnelle (structure spatiale) des mares, dont seule une 

étude partielle portant sur un seul cycle avait été déjà 

présentée (de Bélair, 2004). Ces deux dernières questions 

sont souvent soulevées par les écologues à partir de 

recherches en hydrobiologie, sur la flore comme sur la 

faune (Degiorgi & Grandmottet, 1993 ; Bornette et al., 

1994 ; Gaertner et al., 1998). Cette approche dynamique 

se base sur les analyses statistiques multivariées qui ont 

déjà fait leur preuve en milieu aquatique (Dolédec & 

Chessel, 1987, 1989 ; Beffy & Dolédec, 1991). Une 

double analyse, interclasses (mois et stations), des 

données floristiques permet de séparer l’effet temporel 

de l’effet spatial et permet de discriminer les facteurs 

responsables de l’évolution de la végétation de ces mares 

temporaires de Numidie. 

MATÉRIELS ET MÉTHODES 

Zone d’étude 

La Numidie forme une unité biogéographique précise, 

développée en croissant autour du Djebel Edough 

(sommet : 1 008 m) à l’Ouest d’Annaba (fig. 1). Elle est 

délimitée au nord par la Méditerranée, au sud par un 

ensemble de collines d’altitude moyenne (massifs de la 

Medjerda et de Guelma), n’excédant pas les 1 200 m 

(Djebel Rhorra, à la frontière tunisienne), à l’est par la 

frontière algéro-tunisiennne, correspondant au « rebroussement 

» de l’Atlas tellien vers la mer (Joleaud, 1936) et 

à l’ouest par le massif de Filfila. 

À l’Est comme à l’Ouest de ce massif, se sont développés, 

au Quaternaire récent, deux ensembles dunaires, 

pouvant atteindre 100 m et formant barrage pour les 

eaux d’amont. Aussi, les apports hydriques abondants 

des oueds, rejoignant les plaines sub-littorales numidiennes 

de faible altitude (1 à 5 m), atteignent difficilement 

la mer et forment de nombreux méandres ; c’est le cas, à 

l’est, des oueds El Kébir Est et Bou Namoussa, drainant 

tous les oueds secondaires sur près de 50 km par un 

seul exutoire (oued Mafragh) et Seybouse, tandis qu’à 

l’ouest, un seul oued, El Kébir Ouest, draine toutes les 

eaux d’amont. Ces plaines, où l’écoulement est difficile, 

sont donc facilement inondables et favorisent la formation 

de nombreux marais (garâas) et mares de surface variable. 

De plus, un ensemble de phénomènes tectoniques 

explique la présence d’un grand marais, d’une lagune et 

de deux lacs ; de plus, toutes les plaines sublittorales sont 

soumises à un phénomène de subsidence (Marre, 1992). 

Les dunes elles-mêmes recèlent de très nombreux hydrosystèmes 

plus ou moins éphémères, liées aux ondulations 

du massif dunaire et alimentées par la remontée hivernale 

de la nappe hydrique. 

Cette géomorphologie contrastée détermine un vaste 

éventail d’hydrosystèmes endoréiques, dont l’alimentation 


DYNAMIQUE DE LA VÉGÉTATION DE MARES TEMPORAIRES EN AFRIQUE DU NORD (NUMIDIE ORIENTALE) ◆ 

Figure 1. Carte de situation de 26 mares temporaires 

étudiées en Numidie orientale, identifiées 

selon leur numéro en gras. 

85 

hydrique est assurée temporairement par les pluies, 

les inondations ou la nappe hydrique. Ces mares 

temporaires peuvent être groupées en quatre types, en 

fonction de leur environnement géomorphologique : 

dunes, plaines alluvionnaires, plaines colluvionnaires et 

collines gréseuses (tableau 1 : paramètres mésologiques 

et annexes). Les mares étudiées sont réparties le long de 

deux axes routiers, d’Ouest en Est depuis Annaba : la 

RN 44 au Sud et le CW 109 au Nord. Le premier axe 

traverse les plaines sublittorales, le second longe le massif 

dunaire de Boutelja (fig. 1). 

Échantillonnages 

Durant chaque hydropériode (tableau 2), la végétation 

de 26 mares a été systématiquement échantillonnée chaque 

mois lors de trois cycles, répartis entre les années 1998- 

1999, 1999-2000 et 2000-2001. Ont été intégrés, dans 

la mesure du possible, des relevés précédant la période 

d’inondation ou lui succédant. Il s’avère ainsi possible 

de décrire la végétation résistant à l’assèchement, sinon 

même une flore particulière précédant l’hydropériode ou 

lui succédant. Deux cas ont été distingués : 

— Toutes les hydrophytes, hygrophytes et amphiphytes 

sont affectées d’un indice d’abondance-recouvrement 

de 1 à 10. La somme de ces indices pour une mare 

à une date donnée peut excéder 10 en raison de la 

présence de plusieurs strates : 

* hydriques : hydrophytes flottantes, enracinées, submergées, 

hygrophytes au milieu desquelles se développent 

des hydrophytes, etc., occupant donc des lames 

d’eau différentes ; 

* aériennes : en raison de l’importance de l’ombre dans 

les paramètres écologiques, il a été tenu compte de la 

frondaison des arbres ou arbustes dominant partiellement 

ou totalement certaines mares (cf. p. 17, Frênes, 



86 

Tableau 1. Paramètres mésologiques de vingt-six mares (analyses effectuées par la Tour du Valat). 

Table 1. Mesological parameters of twenty-six ponds (analyses carried out by Tour du Valat). 

N° Mares Altitude Surface Profondeur Argile Limon Sable Sable pH Unité 

Analyses (m) (ares) (cm) (%) fin grossier morphologique 

1 Fei1 3 0.5 120 24.9 15.9 22.6 15.5 4.70 plaine alluvionnaire 




5 Fren 18 7.0 75 45.3 7.5 2.3 0.2 7.90 plaine alluvionnaire 

6 Mess 28 4.5 80 34.3 9.0 17.3 27.0 4.80 plaine colluvionnaire 

7 Gau1 35 10.0 40 21.0 18.3 24.7 18.1 5.60 plaine colluvionnaire 




11 Fedj 110 6.5 100 19.9 10.8 29.9 26.8 5.00 colline gréseuse 

12 Gera 15 3.5 50 22.1 4.5 5.6 56.7 6.25 dune 

13 Isoe 20 3.0 25 14.4 3.3 6.2 66.9 5.70 dune 

14 Ecol 15 3.5 42 23.5 6.6 8.3 44.7 5.35 dune 

15 Sud 8 5.5 70 6.0 1.2 5.2 85.8 6.05 dune 

16 Hrib 10 6.0 130 5.8 0.5 3.6 88.1 4.20 dune 

17 Tama 28 3.5 120 11.3 2.0 19.5 63.6 6.70 dune 

18 Carr 18 3.5 100 4.3 0.7 4.8 88.7 5.95 dune 

19 Mafr 2 6.5 86 6.1 0.7 5.3 87.4 8.35 dune 

20 Sang 3 7.0 86 6.3 0.5 7.0 84.7 8.25 dune 

21 Bouk 1 20.0 90 20.0 14.6 25.1 24.1 7.75 plaine alluvionnaire 

22 Sali 3 100.0 45 56.0 5.9 3.7 0.4 8.15 plaine alluvionnaire 

23 Rupp 2 15.0 70 4.3 0.5 2.2 92.5 7.45 dune 

24 Frin 30 2.0 25 5.1 0.5 7.3 85.9 5.60 rives lac: dune 

25 Bleu 8 0.5 90 5.1 1.3 5.5 86.8 5.45 dune 

26 Buto 9 15.0 70 61.3 1.7 1.1 2.3 5.80 plaine alluvionnaire 



Tableau 2. Hydropériodes de 26 mares temporaires durant 3 cycles. 

Table 2. Hydroperiods of twenty-six ponds for three cycles. 

1998-99 

Sept. Oct. Nov. Décembre Janvier Février Mars Avril Mai Juin Juillet 

1,7,9,16, 7,14,21, 3,12,22, 15,17,19, 12,23, 5,12, 10,17, 1,7, 5.VII 

r: relevé. 

23.XI.98 26,27.XII. 25,28.I.99 23.II. 31.III. 19.IV. 22.V. 12.VI. 

Pluviosité (Salines) 63 61 237 64 157 103 48 44 38 9 13 

Pluviosité (El Kala) 54.1 90 241.2 95.2 

N° Mares 

1 Fe1 r r r r r r r sec 




5 Frê r r r r r r sec 

6 Mes r r r r r r r r 

7 Ga1 r r r r r r r r sec 




11 Fedj r r r r r r r r r (faiblt en eau) 

12 Gér r r r r r r r r 

13 Iso r r r r r r r 

14 BeEc r r r r r r r r 

15 BeS r r r r r r r 

16 HrN r r r r r r r 

17 Tam r r r r r r r r 

18 Carr r r r r r r r r 

19 Maf r r r r r r r 

20 San r r r r r r 

21 Bouk r r r r r r r faiblt en eau 

22 Sal r r r r r r sec 

23 Rup r r r r r r r r r 

24 ElFr r r r 

25 MLB r r r r r r r r r 

26 But r r r r r r r r 

87 



88 

Tableau 2. Hydropériodes de 26 mares temporaires durant 3 cycles (suite). 


1999-2000 

3,13, 10,24, 4,14,21,II. 13,20.III. 10,24.IV. 8,15,29.V. 519,VI 

20.XII.1999 I.2000 

N° Mares 

1 Fe1 à sec: r r r r r r 




5 Frê r r r r r à sec 

6 Mes r r r r r quasi-sec:r r 

7 Ga1 r r r r r quasi-sec:r r 




11 Fedj r r r r r r r 

12 Gér r r r r r sec:r 

13 Iso r r r r r sec:r 

14 BeEc r r r r r r 

15 BeS r r r r r lab: r 

16 HrN r r r r r réd: r 

17 Tam r r r r r r r 

18 Carr r r r r r r 3 bassins 

19 Maf r r r r r r r 

20 San r r r r r r r 

21 Bouk r r r r r r r 

22 Sal r r r r r 3°bas: r 

23 Rup r r r r r r r 

24 ElFr r r à sec:r 

25 MLB r r r r r r 

26 But r r r r r r sec 



Tableau 2. Hydropériodes de 26 mares temporaires durant 3 cycles (suite). 


2000-2001 

19 105 37 108 112 75 19 40 28 0.4 0.4 

Pluviosité 

(Salines) 

33.8 118.5 49.9 85.2 159.2 79.5 21.9 57.8 60.9 Nt Nt 

Pluviosité 

(El Kala) 

26,27.X. 2000 12,24,27.XI. 22,30.XII. 22,26.I.2001 19,25.II-2.II. 23.III-6.IV. 27,29.IV. 25.V. 22,25.VI 

N° Mares 

1 Fe1 sec sec r r r r r sec:r 



12 Gér lab r r r 

13 Iso r (sec de nouveau) r r r r r à sec:r 

14 BeEc r r r r r r r sec/lab:r 

15 BeS r r r r r r 

16 HrN r r r r réd:r quasi-sec:r 

17 Tam r r r r r r r r r 

18 Carr r r r r r r r r 

19 Maf r r r r r r r r r 

20 San r r r r r r r r r 

21 Bouk r r r r r 

22 Sal r r r r r r r 

23 Rup r r r r r r r r r 

24 ElFr r r r 

25 MLB r r r r r r r r r 

26 But r r r r r r r 

89 



90 

Tableau 3. Liste de 136 espèces végétales, inventoriées dans 26 mares de Numidie. 

Table 3. Checklist of 136 plant species, sampled in twenty-six ponds of Numidia (Northeastern Algeria). 

Code Taxa analysés Taxa 

N° 

analyse 

Code Taxa analysés Taxa 

N° 

analyse 

(Quezel & Santa, 1962-1963) (selon Valdès et al., 2002) (Quezel & Santa, 1962-1963) (selon Valdès et al., 2002) 

1 Agsv Agrostis semi-verticillata Polypogon viridis 69 Lyeu Lycopus europeus 

2 Albu Alopecurus bulbosus subsp. macrostachys 70 Lyhy Lythrum hyssopifolia 

3 Alpa Alisma plantago-aquatica 71 Lyju Lythrum junceum 

4 Altr Allium triquetrum 72 Lynu Lythrum numulariaefolia 

5 Apcr Apium crassipes 73 Lysa Lythrum salicaria 

6 Apno Apium nodiflorum 74 Mepu Mentha pulegium 

7 Atlit Atriplex littoralis 75 Mero Mentha rotundifolia 

8 Atpa Atriplex patula A. littoralis ? 76 Myal Myriophyllum alterniflorum 

9 Atpo Atriplex portulacoides Halimione portulacoides 77 Naof Nasturtium officinale 

10 Bere Bellis repens 78 Oefi Oenanthe fistulosa 

11 Bevu Beta vulgaris subsp. maritima Beta maritima 79 Oegl Oenanthe globulosa 

12 Buum Butomus umbellatus 80 Oleu Olea europea 

13 Cadi Carex divisa 81 Orpr Ormenis praecox 

14 Calpe Callitriche pedunculata C. brutia 82 Padi Paspalum distichum 

15 Caob Callitriche obtusangula 83 Pare Panicum repens 

16 Capu Carex punctata 84 Phau Phragmites australis 

17 Catr Callitriche truncata 85 Plco Plantago coronopus 

18 Cede Ceratophyllum demersum 86 Plcr Plantago crassifolia 

19 Char Characeae 87 Poal Populus alba 

20 Chlo Chlorophyceae 88 Poan Poa annua 

21 Chmy Chrysanthemum myconis 89 Poatr Poa trivialis 

22 Cicfi Cicendia filiformis 90 Pola Polygonum lapathifolium 

23 Coco Cotula coronopifolia 91 Poma Polypogon maritimum subsp. subspathaceum 

24 Coli Corrigiola littoralis 92 Pomo Polypogon monspeliensis 

25 Craox Crataegus oxyacantha subsp. monogyna 93 Ponod Potamogeton nodosus 

26 Cyda Cynodon dactylon 94 Pope Potamogeton pectinatus 

27 Cyfl Cyperus flavescens Pycreus flavescens 95 Pore Potentilla reptans 

28 Cylo Cyperus longus subsp. eu-longus 96 Posa Polygonum salicifolium 

29 Cyro Cyperus rotundus subsp. eu-rotundus 97 Potatr Potamogeton trichoides 

30 Daal Damasonium alisma subsp. bourgei 98 Raba Ranunculus baudotii 

31 Disa Digitaria sanguinalis 99 Rafi Ranunculus ficaria 

32 Ecra Echinodorus ranunculoides Baldellia ranunculoides 100 Rahe Ranunculus hederaceus 

33 Elal Elatine alsinastrum 101 Rama Ranunculus macrophyllus 



34 Elbr Elatine brochonii 102 Ramu Ranunculus muricatus 

35 Elemu Eleocharis multicaulis 103 Raop Ranunculus ophioglossifolius 

36 Elhyd Elatine hydropiper 104 Rasa Ranunculus sardous 

37 Elpa Eleocharis palustris 105 Rasc Ranunculus sceleratus 

38 Eupu Euphorbia pubescens 106 Ratr Ranunculus trichophyllus 

39 Fran Fraxinus angustifolia 107 Ruco Rumex conglomeratus 

40 Frlae Frankenia laevis 108 Ruma Ruppia maritima 

41 Fupu Fuirena pubescens 109 Rupu Rumex pulcher 

42 Gapa Galium palustre 110 Ruul Rubus ulmifolius 

43 Glfl Glyceria fluitans 111 Saat Salix atrocinerea 

44 Gllot Glinus lotoides 112 Saleu Salicornia arabica (S. fruticosa) Sarcocornia fruticosa 

45 Hola Holcus lanatus 113 Saso Salsola soda 

46 Ilve Illecebrum verticillatum 114 Sava Samolus valerandi 

47 Irps Iris pseudo-acorus 115 Scce Scirpus cernuus Isolepis cernua 

48 Ishi Isoetes histrix 116 Scho Scirpus holoschoenus Scirpoides holoschoenus 

49 Isve Isoetes velata subsp. typica 117 Scin Scirpus inclinatus 

50 Juac Juncus acutus 118 Scla Scirpus lacustris Schoenoplectus lacustris 

51 Juan Juncus anceps 119 Scma Scirpus maritimus Bolboschoenus maritimus 

52 Juart Juncus articulatus 120 Scni Schoenus nigricans 

53 Jubu Juncus bufonius 121 Seja Senecio jacobaea 

54 Juca Juncus capitatus 122 Sicr Silene coeli-rosa 

55 Jucon Juncus conglomeratus 123 Soni Solanum nigrum 

56 Juef Juncus effusus 124 Sper Sparganium erectum subsp. polyedrum 

57 Juhe Juncus heterophyllus 125 Spfl Spergula flaccida Spergula fallax 

58 Juin Juncus inflexus 126 Spsa Spergularia salina Spergularia marina 

59 Juma Juncus maritimus J. rigidus 127 Taga Tamarix gallica 

60 Jupy Juncus pygmaeus 128 Trfi Trifolium filiforme 

61 Jusu Juncus subulatus 129 Trrep Trifolium resupinatum 

62 Jute Juncus tenageia 130 Trres Trifolium repens 

63 Labi Laurentia bicolor 131 Tyan Typha angustifolia subsp. australis 

64 Lehex Leersia hexandra Oryza hexandra ? 132 Utex Utricularia exoleta 

65 Lemi Lemna minor 133 Utvu Utricularia vulgaris subsp. major 

66 Lino Lippia nodiflora Phyla nodiflora 134 Veaa Veronica anagallis-aquatica 

67 Loco Lotus corniculatus subsp. decumbens 135 Woar Wolffia arrhiza 

68 Lupa Ludwigia palustris 136 Zapa Zanichellia palustris subsp. pedunculata 

91 



92 

entièrement couvert par cette espèce, El Feïd 1, partiellement 

couvert par l’olivier ou, p. 18, Fedjouj, par 

le peuplier et le saule, etc.). 

— Toutes les mésophytes ou espèces ripicoles présentes 

sont notées 1, en raison de leur présence constante et 

de l’information qu’elles apportent sur la formation 

végétale à laquelle appartiennent les mares. 

Analyse inter-classes 

Les données sont par la suite cumulées dans un 

tableau unique, comprenant la somme des relevés des 

trois campagnes (1998-2001), soient 539 relevés et 136 

espèces végétales. Un tel tableau a été soumis à l’analyse 

inter-classes qui conduit à une description de la structure 

globale sur trois cycles. La liste des espèces et leurs codes 

sont rassemblés dans le tableau 3. 

Une première analyse par Analyse en composantes 

principales (ACP) est nécessaire pour accéder à la 

suivante. Sur la base de cette ACP, il est alors possible 

d’effectuer l’analyse inter-classes, dont le déroulement est 

décrit dans le logiciel ADE-4 (Thioulouse et al., 1996). 

RÉSULTATS 

Conséquences écologiques 

globales des hydropériodes 

Les résultats hydrologiques montrent clairement que 

les trois cycles de mise en eau/assèchement sont très 

différents (tableau 2) : 8 mois pour le premier cycle, à 

peine 7 mois pour le deuxième, et 9 mois pour le dernier. 

Ceci reflète parfaitement la variabilité inter-annuelle des 

conditions climatiques en Afrique du Nord. Lorsque 

l’hydropériode est courte, certaines espèces végétales 

ne s’expriment pas, particulièrement en fin de cycle 

qui est brutalement interrompu par la période estivale ; 

c’est le cas par exemple des petites Gentianaceae comme 

Exaculum pusillum ou Cicendia filiformis. Par contre, 

les cycles des espèces principales (cf. analyses : ACP et 

Analyse inter-classes) comme Callitriche obtusangula, 

C. truncata et Ranunculus baudotii ont tendance à se 

chevaucher. 

Si la majorité des mares se remet en eau simultanément, 

un certain nombre d’entre elles accusent un retard 

plus ou moins important (cas des mares El Feïd) ou 

elles achèvent brusquement leur cycle en raison de 

perturbations externes tels que les labours et la mise en 

culture. Enfin, certaines mares peuvent se maintenir en 

eau au-delà de la période d’échantillonnage et présenter 

ainsi une hydropériode qui franchit la période estivale si 

la pluviosité a été bien répartie tout au long de l’année 

(ex. Fedjouj, Tamaris, Mafragh ou Sangliers). Aussi les 

statuts des mares, à l’intérieur d’une même campagne, 

sont-ils variables : aux deux extrêmes, existent des 

mares temporaires très éphémères (hydropériode de 2 à 

3 mois), comme El Frin, et des mares temporaires semipermanentes 

où l’hydropériode peut, une année sur deux 

ou trois, franchir la période estivale. Seules les mares les 

plus profondes appartiennent à ce dernier ensemble. 

Analyse inter-classes 

L’inventaire des 26 mares étudiées et la réalisation de 

539 relevés ont permis d’identifier 136 espèces végétales 

en Numidie, relevés qui ont été soumis à une ACP puis 

à une analyse inter-classes. Les quatre premiers facteurs 

de l’ACP (0.13, 0.09, 0.07 et 0.06) ne mettent pas en 

évidence une structure très forte des données. L’analyse 

inter-classes, concernant la succession des espèces dans 

le temps et la typologie des mares, a été précédée de deux 

tests de permutation, afin de déterminer la pertinence 

d’une telle analyse. Les tests sont dans les deux cas très 

significatifs. 

Structure temporelle (fig. 2) 

L’analyse interclasses affiche deux axes rassemblant 

la majorité de l’information (respectivement, 0.42 et 

0.40). 

L’axe 1 met en évidence dans la structure principale, 

deux groupes (2 semestres ?) : 

— groupe 1 (automne-hiver surtout), marqué par la 

présence de Panicum repens (83), Digitaria sanguinalis 

(31), Isoetes histrix (48) et Lippia nodiflora (= Phyla 

nodiflora) (66) : mois d’octobre à décembre et mois 

de juin (1 à 3 et 9 dans la fig. 2) ; 

— groupe 2 (printemps-été), dominé par Ranunculus 

baudotii (98), Alisma plantago-aquatica (3), Apium 

crassipes (5), Myriophyllum alterniflorum (76), Poa 

annua (88), Ranunculus ophioglossifolius (103), 

R. sardous (104), Scirpus cernuus (115) et Scirpus 

maritimus (119) : mois de janvier à mai (4 à 7 dans 

la fig. 2). 

L’axe 2 distinguerait, dans la structure secondaire, 

deux nouveaux groupes, chevauchant chacun des 

groupes précédents ; il permet la séparation : 

— des mois dominés par Ranunculus baudotii et 

Callitriche obtusangula (15 : en italiques dans la fig. 2) : 



Figure 2. Structure temporelle 

de la végétation des 26 mares 

temporaires. Carte factorielle 

1-2 de l’analyse interclasses 

: 9 mois x 136 espèces 

végétales x 3 cycles. A droite, 

graphe des valeurs propres. 

Chaque mois, d’octobre à 

juin (chiffres romains : X à 

XII et I à VI), est figuré au 

centre (chiffres arabes : 1 à 9, 

entourés d’un cercle). Chaque 

mois est au barycentre des 

relevés (figurés à l’extrémité 

des flèches par un carré noir), 

effectués sur 26 mares durant 

3 cycles. La succession des 

espèces est présentée sous 

forme de quadrats, à échelle 

réduite de la carte factorielle 

initiale. La distribution de 

chaque espèce est figurée par 

des carrés blancs, lorsqu’elle 

est absente, par des cercles, 

lorsqu’elle est présente ; le 

diamètre de ces cercles est 

proportionnel à l’indice 

d’abondance. En caractères 

normaux, les espèces 

dominantes, en italiques, les 

espèces secondaires. 

93 

mois 2 et 3 (novembre-décembre) dans le groupe 1, 

mois 4 et 5 (janvier-février) dans le groupe 2 ; 

— des mois où se développent Apium crassipes, 

des Characeae (19) et Lythrum hyssopifolia (70), 

accompagnés de Cyperus longus (28), Baldellia 

ranunculoides (32), Galium palustre (42) et Paspalum 

distichum (82) : mois 1 et 9 (octobre et juin) d’une 

part, mois 6 à 8 (mars à mai) d’autre part. 

Pouvons-nous en inférer que quatre saisons avec 

des nuances sont mises en évidence dans ce dernier 

découpage ? De plus, cette discrimination apporte-t-elle 

un élément nouveau à la réflexion ? 

Structure spatiale (fig. 3) 

L’analyse inter-classes repose sur deux axes qui 

récoltent une information nettement plus faible que dans 

l’analyse précédente (0.22 et 0.16) ; les autres axes seront 

abordés, mais ils n’apportent qu’une information limitée 

(axes 3 et 4 : 0.10 et 0.09). 

L’axe 1 fait apparaître une structure principale, 

répartissant les mares en deux groupes : 

— groupe 1 à Characeae (19) et Juncus acutus (50) 

qu’accompagnent Juncus maritimus (J. rigidus Valdès 

et al., 2002) (59), Typha angustifolia (131) et Tamarix 

gallica (127) : mares 3 à 6, 11 et 18 à 26. 



94 

— groupe 2 à Callitriche obtusangula (15), Isoetes velata 

(49) et Myriophyllum alterniflorum (76) principalement 

et, secondairement, Baldellia ranunculoides (32), 

Eleocharis palustris (37) et Glyceria fluitans (43) : 

mares 1, 2, 7 à 10 et 12 à 17. 

L’axe 2 fait ressortir une structure secondaire distinguant 

dans chacun des groupes un nouveau groupement, défini 

par Scirpus maritimus (= Schoenoplectus maritimus) (119 : 

en italiques dans la fig. 3) et Alisma plantago-aquatica (3 : 

en italiques), qu’accompagne souvent Scirpus lacustris (= 

Schoenoplectus lacustris) (118). 

Cette analyse permet une partition des 26 mares de 

la manière suivante : 

— du groupe 1, se dégage un groupe à Scirpus maritimus, 

S. lacustris et Typha angustifolia : mares 1, 2, 4, 

15, 16 et 17, 

— du groupe 2 se détache, de la même manière, un 

groupe à Scirpus maritimus : mares 3, 4, 5, 6, 21, 22, 

24, 25 et 26. 

En résumé, il existe trois groupes nettement éloignés 

des axes 1 et 2, et donc clairement identifiables par la 

dominance des espèces qu’ils contiennent : 

* groupe 1 à Characeae, J. maritimus et J. acutus : 11, 

18 à 20 et 23 ; 

* groupe 2 à Callitriche obtusangula, Isoetes velata et 

Myriophyllum alterniflorum : 7 à 10 (8 masqué par 9) 

et 12 à 14 ; 

* groupe 3 à Callitriche obtusangula, Scirpus maritimus 

et Alisma plantago-aquatica : 1, 2 (masqué par 4), 15, 

d’une part, 3 à 6 avec 21 et 26 d’autre part. 

En outre, un groupe mal défini par ces axes et 

rassemblé vers l’origine, comporte les mares 16, 17 

Figure 3. Structure spatiale 

de la végétation des 26 mares 

temporaires. Carte factorielle 1- 

2 de l’analyse inter-classes : 26 

mares x 136 espèces végétales x 3 

cycles. En haut, à droite, graphe 

des valeurs propres. Chaque 

station (affectée d’un chiffre de 

1 à 26) est au barycentre de 

l’ensemble des relevés (liés à la 

durée de l’hydropériode) effectués 

durant 3 cycles. La distribution 

de chaque taxon est représenté 

dans un quadrat à échelle 

réduite de la carte factorielle 

principale. La distribution de 

chaque espèce est figurée par 

des carrés blancs, lorsqu’elle est 

absente, par des cercles, lorsqu’elle 

est présente ; le diamètre de ces 

cercles est proportionnel à l’indice 

d’abondance. La typologie des 

mares est définie par 3 groupes 

végétaux, clairement identifiés. 



(masqué par 25), 22, 24 et 25. Chacune de ces mares 

est originale : 

(i) soit par sa composition floristique (ainsi, les mares 

16 et 17 relèvent à la fois des groupes 2 et 3 ; la mare 22, 

marquée par une forte salinité, comporte une association 

végétale particulière à Salicornia arabica et Salsola soda ; 

la station 25 est dominée par plusieurs espèces, absentes 

des autres, comme Wolffia arrhiza, Ludwigia palustris et 

Oryza hexandra) ; 

(ii) soit par la brièveté de son hydropériode ; tel est le 

cas de la station 24. 

Les axes 3 et 4 ne prennent en charge que des 

situations particulières en raison de la dominance d’une 

espèce. Ainsi, Wolffia arrhiza (135) est permanente et 

abondante dans la mare du lac Bleu et Ruppia maritima 

(108) dans la mare Ruppia. 

DISCUSSION 

Structure temporelle : 

succession des espèces végétales 

A propos du climat en Algérie, Seltzer (1946) souligne : 

« Le climat est partout méditerranéen, i.e. caractérisé par une 

saison pluvieuse allant en moyenne de septembre à mai, et 

par un été sec et ensoleillé ». La végétation exprimée lors 

des mois d’été est dans notre étude, par définition, sousreprésentée 

dans notre échantillonnage. La structure 

bi-partite recouvre donc surtout la « saison pluvieuse » ; 

elle est déterminée par l’axe 1 et sépare deux ensembles 

végétaux, l’un englobant les mois d’octobre à décembre et 

de juin, définis par Panicum repens et ses associés, l’autre, 

les mois de janvier à mai, marqués par la présence de 

Ranunculus baudotii et ses associés. Le premier ensemble 

se développe en effet particulièrement au début de 

l’été (juillet et août n’ont pas été échantillonnés) et se 

poursuit durant l’automne, où les Graminées, comme 

Panicum repens et Paspalum distichum à développement 

concomitant, entrent dans le stade végétatif. Lors de la 

remise en eau des mares, un délai de germination et de 

floraison est évidemment nécessaire pour que R. baudotii 

puisse, à son tour, coloniser les mares. 

L’axe 2 introduit des nuances, et à R. baudotii 

s’adjoint la présence de Callitriche obtusangula. Cette 

dernière espèce précède la mise en place de R. baudotii 

et peut persister longtemps en association avec elle ; elle 

disparaît lorsque R. baudotii occupe progressivement les 

mares et entre en compétition avec C. obtusangula pour 

la lumière. Elle est capable de germer, se développer et 

occuper la niche écologique, libérée de toute autre espèce, 

dès le début de l’hydropériode des mares. À la lecture 

de la fig. 3, c’est la succession habituelle des groupes 

écologiques, sinon des espèces, « imbriquées en écailles » 

(Godron, 1967) que laissent apparaître les distributions 

de C. obtusangula (stades 3, 4, 5 et 6), R. baudotii (stades 

5, 6 et 7), Alisma plantago-aquatica (stades 5, 6 7 et 8), 

Apium crassipes (stades 6, 7, 8 et 9) et les Characeae 

(stades 8, 9, 1 et 2). 

Il est important de noter que les stades 1 à 4 sont 

condensés, tandis que les suivants sont plus dispersés. 

Les premiers stades ne mettent que peu en évidence 

les différences entre stations, où dominent partout les 

Graminées à floraison estivalo-automnale, Panicum repens 

et Paspalum distichum. Durant les stades suivants, la flore 

se diversifie (cf. typologie) et son analyse discrimine plus 

clairement les étapes de la succession des espèces. 

Structure spatiale : typologie des stations 

Les trois groupes (1, 2, 3) signalés ci-dessus ont été 

décrits par les phytosociologues comme appartenant à 

trois ordres différents (Braun-Blanquet et al., 1951). 

Le groupe 1 à Juncus acutus, J. maritimus et Typha 

angustifolia, avec les nuances qui s’imposent pour le 

Sud méditerranéen et particulièrement la Numidie, 

formerait une association appartenant aux Juncetalia 

maritimi Br.-Bl. 1931 (Gehu, 1993) ; les groupements 

de cet ordre appartiendraient aux « prés salés », 

selon ces auteurs, et seraient surtout constitués de 

Glumiflores. Ce groupement à Juncus acutus et 

J. maritimus a fait l’objet de nombreux travaux, mais 

les Characeae ne sont pas incluses dans la définition 

de ce syntaxon. Pourtant, Feldmann (1946) a 

décrit un grand nombre de Characeae en Numidie : 

Tolypella mucronata, huit espèces de Nitella et cinq 

Chara, sur 35 espèces en Afrique du Nord, dont 28 

pour l’Algérie. Les Characeae rencontrées dans les 

mares étudiées appartiennent à divers milieux, dont 

certains plus ou moins saumâtres. Une actualisation 

de ces travaux semble donc indispensable et devrait 

ouvrir à une identification des Characeae récoltées. 

L’association entre plantes pérennes (hélophytes ou 

succulentes), des genres Arthrocnemum, Juncus et 

Scirpus est connue et a fait l’objet de travaux, comme 

ceux de l’équipe d’Espinar et al. (2002) sur la zonation 

des macrophytes submergées dans les marais salants 

méditerranéens. 

95 



96 

Le groupe 2, souvent dénommé “prairies à Isoetes”, a 

particulièrement été étudié par Chevassut & Quézel 

(1956), Poirion & Barbero (1965), Aubert & Loisel 

(1971), Barbero (1965, 1967), Rhazi et al. (2004). 

La présence importante de formations tourbeuses 

dans la région est favorable à la formation de ces 

prairies humides à très humides. Souvent situées en 

interface entre les dunes et les plaines sublittorales, 

donc à biodiversité souvent très élevée, ce type de 

formation appartiendrait à l’ordre Isoetalia. Br.-Bl. 

1931 (Gehu, 1993). De manière générale, les mares 

de Numidie comportant Isoetes velata (probablement, 

subsp. typica) sont situées dans des prairies mésophiles 

à Isoetes histrix ; la texture est sableuse avec des sables 

grossiers de dunes, et des sables fins ayant pour origine 

les grès et argiles dits de Numidie (Joleaud, 1936). 

Le groupement à Isoetes et Myriophyllum alterniflorum 

est, sans conteste, celui qui attire le plus notre attention sur 

le plan de la richesse spécifique, composition floristique 

et valeur patrimoniale en raison des espèces présentes, 

souvent rares ou appartenant à l’élément afrotropical. 

Médail et al. (1998) pour la France méditerranéenne, 

Quézel (1998) et Grillas et al. (2004) pour les pays 

méditerranéens mettent en exergue ce groupement et les 

espèces qui le composent, y compris certaines Characeae, 

qui lui sont associées. Ainsi, Callitriche truncata, 

Ceratophyllum demersum, Elatine alsinastrum, E. brochoni, 

Illecebrum verticillatum, Cicendia filiformis, Exaculum 

pusillum (non noté dans la liste des espèces, mais présent 

dans les stations Gauthier) sont considérées comme rares 

en Algérie, et spéciales à la Numidie. Trois espèces sont 

d’origine afrotropicale : Scirpus inclinatus, Utricularia 

exoleta et Wolffia arrhiza qui peut être envahissante. Une 

endémique algéro-tunisienne, Bellis repens, est largement 

présente dans ce groupement. D’autres espèces, connues 

sur tout le pourtour de la Méditerranée, appartiennent 

également à ce groupement, comme Apium crassipes 

ou Ranunculus sceleratus, considérées comme rares 

en Algérie. R. ophioglossifolius, Baldellia ranunculoides 

ou Glyceria fluitans sont, elles, communes. Les mares 

temporaires, occupées par ce groupement, présentent la 

plus grande richesse spécifique et elles devraient avoir 

un statut prioritaire de protection. Or, la plupart d’entre 

elles sont convoitées en raison d’une autre richesse, celle 

de leur sol, dont la texture recèle un taux élevé de matière 

organique, favorable à certaines cultures (arachide, 

notamment). Ce groupement avait fait l’objet en Algérie 

de plusieurs études, citées plus haut, mais il n’avait pas 

été décrit pour la Numidie. 

Le groupe 3 à Scirpus maritimus dépendrait de l’ordre 

Phragmitetalia, et plus particulièrement de l’association 

Scirpetum maritimi compacti Dahl et Hadac 1941 

(Gehu, 1993). La notation des auteurs est éclairante : 

« l’association colonise les vases fines au bord des étangs et 

des fleuves de la région côtière ; elle est faiblement halophile 

et supporte un dessèchement temporaire ». 

Ce groupement à Scirpus maritimus a été largement 

étudié, et l’espèce principale, qui occupe de grandes surfaces 

autour de la Méditerranée, a fait l’objet de plusieurs 

investigations récentes. Ainsi, Charpentier et al. (2000) 

et Charpentier (2002) ont étudié les conséquences de la 

croissance clonale sur la dynamique et la structure spatiale 

des populations de ce scirpe. Il est connu comme 

étant un bioindicateur de l’alcalinité du sol, mais dans la 

région d’étude, l’écologie de cette unité nuancerait cette 

affirmation (cf. annexes). Ce groupement comporte 

souvent des Characeae, qu’il possède en commun avec le 

groupe précédent. Trois espèces liées à ce groupe – Oryza 

hexandra, Fuirena pubescens et Glinus lotoides –appartiennent 

à l’élément tropical. La dernière tapisse le fond des 

mares et des marais, sinon des lacs durant toute la période 

estivale. 

Cette première typologie mérite toutefois quelques 

compléments. En effet, si les espèces les plus caractéristiques 

de ces mares sont souvent identiques à celles 

de l’Europe méditerranéenne, d’autres sont particulières 

à l’Afrique du Nord, notamment à la Numidie. Elles 

forment ainsi des associations locales, étudiées par Gehu 

(1993) qui fait également un état des végétaux rares, sinon 

rarissimes de la région. L’élément le plus original est sans 

conteste la présence significative d’espèces d’origine biogéographique 

afrotropicale (8,8 % des végétaux recensés). 

Ce caractère n’avait pas échappé à Lefranc (1865) et il 

a aussi été signalé dans l’étude générale concernant les 

zones humides de la région d’El Kala par Stevenson et al. 

(1989). La présence de cet élément est plus importante 

dans les deux groupes à Isoetes et Scirpus maritimus, avec 

Fuirena pubescens, Glinus lotoides, Leersia (= Oryza) hexandra, 

Scirpus inclinatus, Utricularia exoleta et Wolffia arrhiza, 

toutes réputées rares au niveau national. Ainsi la Numidie, 

où ces taxons sont souvent signalés comme exclusifs de 

cette région, représente le siège de cette « poche afrotropicale 

». Il faudrait y ajouter deux espèces, très envahissantes 

et de même origine : Panicum repens et Paspalum distichum. 

Cette situation biogéographique suppose des conditions 

écologiques particulières à la Numidie, qui sont favorables 

au maintien et au développement de tels végétaux 

exigeants en température et humidité. 



L’originalité de deux mares (déterminée par les axes 3 

et 4) mérite d’être soulignée. Vincent (2004) invite à 

prendre en compte cette originalité, quand il indique 

que l’« on n’a plus ici une opposition entre le typique (ici une 

typologie à trois groupements) et le singulier (les deux mares, 

l’une à Ruppia maritima, l’autre à Wolffia arrhiza), mais ce 

dernier est présent comme l’obstacle qui permet et exige que 

l’on dépasse la première généralisation ». Ces deux mares 

posent des questionnements écologiques en raison de 

leur composition floristique bien à part. En Numidie, 

aucune de nos investigations ne nous a permis de trouver 

une autre mare dominée ainsi par Ruppia maritima qui 

entre en concurrence avec les autres espèces végétales 

particulièrement pour la lumière. Simultanément, en 

oxygénant le milieu, elle favorise une multitude de niches 

écologiques, occupées par des Amphibiens comme la 

Grenouille verte, par le phytoplancton et le zooplancton. 

Le groupement où intervient Wolffia arrhiza est plus 

fréquent ; il occupe non seulement la mare du lac Bleu, 

mais les rives de certains lacs ou étangs en Numidie, 

chaque fois que le milieu s’eutrophise (Samraoui & de 

Bélair, 1997). 

CONCLUSION 

L’échantillonnage de 26 mares temporaires réalisé 

durant trois cycles (1998-2001) a permis d’inventorier 

136 espèces végétales en Numidie. L’ACP puis l’analyse 

inter-classes ont permis de séparer deux effets : temporel 

et spatial. La succession des espèces (effet temporel) met 

en évidence la séparation de l’année en deux groupes, l’un 

dominé par Panicum repens en automne-été, l’autre par 

Ranunculus baudotii en hiver-printemps ; une partition 

dans ce dernier groupe paraît plus aléatoire en raison 

de la structure habituelle « en écailles » des espèces. La 

typologie met très nettement en évidence trois groupes, 

caractérisés successivement par (i) Juncus maritimus, 

J. acutus et des Characeae, (ii) par Scirpus maritimus et 

ses associés et enfin (iii) par Isoetes velata et Myriophyllum 

alterniflorum. 

Comme l’on dispose d’autres paramètres environnementaux 

caractéristiques du milieu, une analyse supplémentaire 

pourra être effectuée en croisant les tableaux 

floristique et mésologique afin d’établir le bien-fondé de 

cette typologie. L’analyse de co-inertie représente probablement 

l’instrument répondant le mieux à ce type de 

manipulation (Thioulouse et al., 1996). 

L’ensemble de ces hydrosystèmes temporaires correspond 

à une large amplitude du pH allant d’une alcalinité 

élevée à une acidité forte. Ces mares rassemblent une 

flore diversifiée, dont beaucoup d’espèces sont rares à 

très rares en Algérie ; elles sont partiellement d’origine 

tropicale (8 espèces), et majoritairement inféodées à la 

seule Numidie. Cette originalité confère donc aux mares 

temporaires de Numidie un intérêt patrimonial indéniable. 

Face à la vulnérabilité de ces mares, ces résultats 

devraient inciter à la mise en place de rapides mesures 

conservatoires afin de sauver ces fragiles écosystèmes 

d’une disparition sans doute inéluctable liée à l’anthropisation 

croissante du littoral algérien. 

REMERCIEMENTS 

L’auteur tient à remercier le Laboratoire de recherche 

sur les zones humides d’Annaba et la Station biologique de 

la tour du Valat pour les aides appréciables apportées dans 

le cadre de cette étude. Merci également au Dr Frédéric 

Médail (IMEP-CNRS, Univ. Aix-Marseille III) pour ses 

remarques et suggestions. 

Bibliographie 

AUBERT G. & LOISEL R., 1971. Contribution à l’étude des groupements 

des Isoeto-Nanojuncetea et des Helianthemetea annua 

dans le sud-est méditerranéen français. Ann. Univ. Provence, 

Marseille, 45 : 203-241. 

BARBERO M., 1965. Groupements hygrophiles de l’Isoetion dans 

les Maures. Bull. Soc. Bot. Fr., 112 : 276-290. 

BARBERO M., 1967. L’Isoetion des Maures. Groupements 

mésophiles – Étude du milieu. Ann. Fac. Sci. Marseille, 39 : 

25-37. 

BRAUN-BLANQUET J., ROUSSINE N. & NÈGRE R., 1951. Les 

Groupements végétaux de la France méditerranéenne. CNRS, 

Paris, 297 p. 

BEFFY J.-L. & DOLÉDEC S., 1991. Mise en évidence d’une typologie 

spatiale dans le cas d’un fort effet temporel : un exemple 

en hydrobiologie. Bull. Écol., 22 : 3-11. 

BORNETTE G., AMOROS C., CASTELLA C. & BEFFY J.-L., 1994. 

Succession and fluctuation in the aquatic vegetation of two 

formers Rhône River channels. Vegetatio, 110 : 171-184. 

CHARPENTIER A., GRILLAS P. & T HOMPSON J.-D., 2000. The 

effect of population size limitation on fecundity in mosaic 

populations of the clonal macrophyte Scirpus maritimus 

(Cyperaceae). Am. J. Bot., 87 : 502-507. 

CHARPENTIER A., 2002. Consequences of clonal growth for plant 

mating. Evol. Ecol., 15: 521-530. 

CHEVASSUT G. & QUÉZEL P., 1956. Contribution à l’étude des 

groupements végétaux de mares temporaires à Isoetes velata 

97 



98 

et de dépressions humides à Isoetes hystrix en Afrique du 

Nord. Bull. Soc. Hist. Nat. Afrique du Nord, 47 : 59-73. 

DE BÉLAIR G. & BENCHEIKH-LEHOCINE M., 1987. Composition et 

déterminisme de la végétation d’une plaine côtière marécageuse : 

la Mafragh (Annaba, Algérie). Bull. Écol., 18 : 393-407. 

DE BÉLAIR G., 2004. Statut spatio-temporel de 25 mares temporaires, 

cycle 2000-2001 (Numidie orientale, Algérie). 

Conférence internationale : Les mares temporaires méditerranéennes 

: de la connaissance à la gestion. Roquebrune-sur- 

Argens, communication : 6 p. 

DEGIORGI F. & GRANDMOTTET J.-P., 1993. Relations entre la 

topographie aquatique et l’organisation spatiale de l’ichtyofaune 

lacustre : définition des modalités spatiales d’une 

stratégie de prélèvements reproductibles. Bull. Fr. Pêche 

Pisciculture, 329 : 199-220. 

DEIL U., 2005. A review on habitats, plant traits and vegetation 

of ephemeral wetlands – a global perspective. Phytocoenologia, 

35 : 533-705. 

DOLEDEC S. & CHESSEL D., 1987. Rythmes saisonniers et composantes 

stationnelles en milieu aquatique. I. Description 

d’un plan d’observation complet par projection de variables. 

Acta Œcol., Œcol. Gen., 8 : 403-426. 

DOLEDEC S. & CHESSEL D., 1989. Rythmes saisonniers et 

composantes stationnelles en milieu aquatique. II. Prise en 

compte et élimination d’effets dans un tableau faunistique. 

Acta Œcol., Œcol. Gen., 10 : 207-232. 

ESPINAR J.-L., GARCIA L.V., GARCIA MURILLO P. & TOJA J., 

2002. Submerged macrophyte zonation in a Mediterranean 

salt marsh : a facilitation effect from established helophytes 

inquest. J. Veg. Sci., 13 : 831-840. 

FELDMANN G., 1946. Les Charophycées d’Afrique du Nord. 

Bull. Soc. Hist. Nat. Afrique du Nord, 37 : 64-118. 

GAERTNER J.-C., CHESSEL D. & BERTRAND J., 1998. Stability 

of spatial structures of demersal assemblages : a multitable 

approach. Aquat. Living Resour., 11 : 75-85. 

GAUTHIER-LIÈVRE L., 1937. Recherches sur la flore des eaux 

continentales de l’Afrique du Nord. Bull. Soc. Hist. Nat. 

Afrique du Nord. Mém. hors série, Alger. 300 p. 

GEHU J.-M., KAABECHE M. & GHARZOULI R., 1993. Phytosociologie 

et typologie des habitats des rives des lacs de 

la région d’El Kala (Algérie). Coll. Phytosociol. XXII. 

Syntaxonomie typologique des habitats. Bailleul, p. 297-331. 

GODRON M., 1967. Les groupes écologiques imbriqués en 

écailles. Œcol. Plant., 2 : 217-226. 

GRILLAS P., GAUTHIER P., YAVERCOSKI N. & PERENNOU C., 2004. 

Les mares temporaires méditerranéennes. Tour du Valat. Volume I 

– Enjeux de conservation, fonctionnement et gestion : 119 p. ; 

volume II – Fiches espèces : 128 p. 

JOLEAUD L., 1936. Étude géologique de la région de Bône et de 

La Calle. Bull. Serv. Carte géolog. Algérie. Imp. Typo-litho et 

C ie , Alger, 2° série, n° 12, 185 p., 4 pl., 25 fig. et tab. 

LEFRANC E., 1865. La Calle. Topographie, botanique et climatologie. 

Bull. Soc. Bot. Fr., 12 : 415-431. 

MARRE A., 1992. Le Tell oriental algérien. De Collo à la frontière tunisienne. 

Étude géomorphologique. OPU, Alger, 2 vol. : 624 p. 

MARTI C. & ESTEBAN A., 2001. Across The Waters-Newsletter. 

Posidonia. WWF Mediterranean Programme. 3 p. 

MÉDAIL F., MICHAUD H., MOLINA J., PARADIS G. & LOISEL R., 

1998. Conservation de la flore et de la végétation des mares 

temporaires dulçaquicoles et oligotrophes de France méditerranéenne. 

Ecol. Medit., 24 : 119-134. 

MORGAN N.C., 1983. An ecological survey of standing waters 

in North West Africa : II. Site descriptions for Tunisia and 

Algeria. Biol. Conserv., 24 : 83-113. 

POIRION L. & BARBÉRO M., 1965. Groupements à Isoetes velata 

A. Braun (Isoetes variabilis Le Grand). Bull. Soc. Bot. Fr., 

112 : 436-442. 

QUÉZEL P. & SANTA S., 1962-1963. Nouvelle Flore de l’Algérie et 

des régions désertiques méridionales. CNRS. Paris. 1170 p. 

QUÉZEL P., 1978. Analysis of the flora of Mediterranean and 

Saharan Africa. Ann. Missouri Bot. Gard, 65 : 479-534. 

QUÉZEL P., 1998. La végétation des mares transitoires à Isoetes en 

région méditerranéenne, intérêt patrimonial et conservation. 

Ecol. Medit., 24 : 111-117. 

RHAZI L., GRILLAS P., M OUNIROU TOURÉ A. & TAN HAM L., 

2001. Impact of land use and activities on water, sediment 

and vegetation of temporary pools in Morocco. C.R. Acad. 

Sci. Paris. Série III, Sci. Vie, 324 : 165-177. 

RHAZI M., GRILLAS P., CHARPENTIER A. & MÉDAIL F., 2004. 

Experimental management of Mediterranean temporary 

pools for conservation of the rare quillwort Isoetes setacea. 

Biol. Conserv., 118 : 675-684. 

SAMRAOUI B. & G. DE BÉLAIR, 1997. The Guerbes-Senhadja 

wetlands (N.E. Algeria). Part I : an overview. Écologie, 28, 

3 : 233-250. 

SAMRAOUI B. & G. DE BÉLAIR, 1998. Les Zones humides de la 

Numidie orientale. Bilan des connaissances et perspectives de 

gestion. Synthèse. Univ. Annaba, Algérie. 90 p. 

SELTZER P., 1946. Le Climat de l’Algérie. Univ. Alger., Alger, 

219 p. 

STEVENSON A.C., SKINNER J., HOLLIS G.E. & SMART M., 1988. 

The El Kala Park and environs, Algeria : An ecological evalutation. 

Env. Conserv, 15 : 335-348. 

THIOULOUSE J., D. CHESSEL, S. DOLEDEC & J.-M. OLIVIER, 

1996. ADE-4 : a multivariate analysis and graphical display 

software. Statistics and Computing, 7 : 75-83. 

VALDÈS B., REDJALI M., ACHHAL EL KADMIRI A., JURY J.-L. & 

MONTSERRAT J.-M., 2002. Catalogue des Plantes vasculaires 

du Nord du Maroc, incluant des clés d’identification. CSIC, 

Madrid. Vol. I & II, 1 007 p. 

VINCENT H., 2004. Parler des animaux : une rhétorique furtive. 

Le Courrier de l’Environnement INRA, 51 : 55-60. 



ANNEXES 

Paramètres floristiques et mésologiques 

des 26 mares temporaires étudiées 

D’ouest en est, les 26 mares temporaires décrites sont situées 

dans des environnements différents déterminés par des paramètres 

mésologiques et des formations végétales diverses. Ces mares ont 

été classées selon quatre unités géomorphologiques, et une brève 

description mésologique et floristique est fournie. 

1. Plaines sublittorales alluvionnaires 

* Boukamira (21) est une mare littorale sablo-argileuse, à 

proximité de la ville d’Annaba ; son pH est alcalin : 7.75. Elle sert 

de dépotoir et d’exutoire pour les eaux usées, mais ce n’est pas, du 

moins actuellement, un obstacle à l’expression d’une biodiversité 

floristique élevée. Elle fait partie d’un complexe de mares du même 

type, naturelles ou artificielles. Elle est entourée d’une prairie à 

Trifolium campestre, Melilotus infestus et M. siculus, Lolium rigidum, 

Hordeum marinum, Crypsis alopecuroides, Medicago intertexta, 

Ricinus communis et de quelques espèces rares, comme Heliotropium 

curassavicum, Scirpus littoralis (origine biogéographique : paléosubtropicale) 

ou Rumex algeriensis (endémique). 

* Salines (22) fait partie d’un ensemble, aujourd’hui 

abandonné, de bassins de décantation pour la production de sel ; elle 

est également en position littorale, sans contact direct avec la mer. 

La dominante est argileuse (56 %). Elle comporte une formation à 

Hordeum marinum et Lolium rigidum, accompagnés de Puccinellia 

distans, Melilotus infestus, Medicago intertexta et Rumex algeriensis ; 

cette mare se caractérisée par un milieu alcalin (pH = 8.15) et un 

grand nombre d’espèces messicoles en raison de la présence tout 

autour de terres cultivées. 

* Mares d’El-Feïd (1 à 4). Elles sont situées sur les rives 

sud-ouest du marais de la M’krada ou plaine de la Mafragh 

(de Bélair & Bencheikh-Lehocine, 1987) ; elles ont été creusées 

pour l’abreuvement du bétail dans une prairie d’une grande 

richesse spécifique (ainsi, un transect effectué sur 150 m avait 

mis en évidence près de 150 taxons). Citons Glinus lotoides qui 

se développe dans les mares asséchées en été, Dactylis glomerata, 

Gaudinia fragilis, Myosotis sicula, Hordeum bulbosum, Sonchus asper, 

Stachys arvensis, Medicago intertexta, Asphodelus aestivus, Hypochoeris 

radicata, Festuca elatior et Phalaris arundinacea, ces deux dernières 

souvent dominantes au printemps. La dominante texturale est 

argilo-limoneuse, sauf pour Feïd 1 à texture équilibrée. Malgré la 

proximité d’un marais relativement alcalin, les pH sont acides (de 4 

à 5). Deux Orchidées sont disséminées dans cette prairie : Serapias 

strictiflora et S. parviflora. 

* Frênes (5) est localisé dans la ripisylve à Ulmus campestris et 

Populus alba de l’oued El Kébir-Est. La zone où s’est développée 

cette mare a été plantée de Fraxinus angustifolia. Sa texture 

est principalement argileuse (45.3 %) ; son environnement est 

constitué par une formation à Asphodelus aestivus, Urginea maritima, 

Hypochoeris radicata, Ranunculus bulbosus, Plantago lanceolata, 

Medicago littoralis et Ormenis mixta. 

2. Plaines sublittorales colluvionnaires 

* Messida (6) présente une texture sablo-argileuse (sable : 40 % 

et argile : 34 %) ; située en aval de collines gréseuses et à proximité 

de l’oued Messida, son pH est très acide (4.80). Proche de terres 

cultivées, Medicago sp., Melilotus sp. et Trifolium sp. abondent avec 

quelques graminées. 

* Les mares Gauthier (7 à 10). Leur dominante texturale 

est plutôt sableuse pour les mares 7 et 8 (21 % d’argile, 18 % de 

limon et 36 % de sable), nettement plus argileuse (38 %) pour les 

mares 9 et 10. Leur pH est également acide (de 4.80 à 5.60). Des 

bosquets de Quercus suber, Pistacia lentiscus et Myrtus communis 

sont dispersés dans une prairie humide, jalonnée de mares aux 

dimensions variables (de quelques ares à plusieurs hectares). La 

formation est dominée, en dehors des vestiges d’une subéraie, par 

Asphodelus aestivus, Hypochoeris radicata, Eryngium barrelieri et, 

au gré des saisons, par Bellis annua, Triglochin laxiflora, Leucojum 

autumnale, Romulea bulbocodium, Briza maxima et Radiola linoides. 

* Butomes (26). Elle occupe une dépression étalée en longueur, 

rapidement inondée lors des crues de l’oued El Kébir-Est. Située 

au nord-est du marais du Mkrada, elle présente une végétation 

similaire. Sa texture est dominée par les argiles (61.3 %). Son 

caractère remarquable réside dans une population d’extension 

variable de Butomus umbellatus ; cette espèce, considérée en Algérie 

comme très rare (Quézel & Santa, 1962), a été signalée aux environs 

d’Alger (Maison-Carrée = El Harrach ?) et surtout en Numidie (5 

stations ont été repérées entre Skikda et El Kala). Scirpus littoralis est 

également présente avec Dipsacus sylvestris et Verbena officinalis. 

3. Massif dunaire 

* El Frin (24) appartient à un ensemble, souvent très fugace, de 

mares mises en eau sur les rives du lac Oubeïra, favorisées par une 

série de micro-reliefs et de micro-dépressions. La totalité de cette 

prairie humide, plus ou moins tourbeuse à l’origine, est désormais 

cultivée depuis les années 1990 (arachide, principalement) ; la 

végétation s’est rapidement banalisée, la tourbe s’étant minéralisée. 

Peu d’espèces ont le temps de se développer du fait de la brève 

hydropériode (3 mois en moyenne) et elles sont rapidement 

remplacées par un tapis d’Ormenis mixta. Signalons que les fossés 

proches sont favorables à certaines Orchidées, comme Serapias 

stenopetala ou Ophrys bombyliflora. 

* La mare Ruppia (23) est adossée en pleine dune aux rochers 

maritimes dans une dépression. Sa texture est évidemment dominée 

par les sables (95 %) avec un pH alcalin (7.45). Ceci explique la 

présence massive de Ruppia maritima, seule mare repérée avec ce 

taxon. Tamarix gallica (T. africana ?) l’entoure presque totalement en 

ceinture, suivie d’une ceinture externe de Phillyrea latifolia, Myrtus 

communis et Quercus coccifera. Le sol est légèrement tourbeux, mais 

fortement dégradé par le piétinement et le pâturage du bétail. 

99 



100 

* Mare Lac Bleu (25). Temporairement alimentée par une 

source se déversant sur le lac, elle est située à mi-pente et sert à 

abreuver le bétail et, éventuellement à irriguer les terres, toutes 

cultivées, sur le pourtour. La texture est composée à près de 93 % 

de sable, mais l’accumulation de matière organique (une petite 

aulnaie s’est développée en aval) crée un milieu plus ou moins 

tourbeux riche en azote (C/N = 5.08). Ces conditions mésologiques 

sont à l’origine du développement des Utricularia. L’anthropisation 

(tendance à l’eutrophie) se traduit par le fort développement de 

Wolffia arrhiza et de Nasturtium officinale. 

* Gérard (12) se situe en aval du cordon littoral septentrional, 

s’étalant d’est en ouest entre Annaba et le Djebel Koursi. Le sable 

domine (62 %), l’argile ne représentant que 22 % de la fraction 

texturale. Adossée à ce cordon, l’accumulation de la végétation 

et donc de la matière organique favorise la formation de tourbe. 

En situation d’interface entre dunes et plaines sublittorales, elle 

présente une très grande biodiversité, où domine la magnocariçaie 

(Carex elata, C. acutiformis, Cladium mariscus, etc.), ainsi que Salix 

atrocinerea, Erica scoparia, Alliaria officinalis, Oxalis corniculata, 

Brassica decumbens et les rares Anagallis crassifolia et Radiola 

linoides. 

* Isoetes (13) et Berrihane école (14) sont en situation de 

prairie humide tourbeuse, à texture sableuse (76 % de sable) sur 

dune et à pH acide (5.6). Le plus souvent mises en culture (arachide 

principalement), ces mares sont très vulnérables et d’une très grande 

richesse spécifique, y compris en espèces patrimoniales avec la 

présence de plusieurs taxons en voie de disparition. Prédominent 

les Isoetes, Isolepis setacea, Imperata cylindrica et Serapias strictiflora. 

Les quelques haies basses qui jalonnent ces prairies rassemblent 

les reliques de forêts humides sur tourbes (aulnaies), avec Erica 

scoparia, Pteridium aquilinum, Myrtus communis, Salix atrocinerea, 

Carex sp., etc. 

* Berrihane Sud (15) se trouve à la sortie ouest du village de 

même nom. C’est une station à dominante sableuse (91 % de sable) 

et à pH acide (6.05). Dépression largement entourée d’Eucalyptus 

sp., ayant servi de déversoir à un élevage de poulets, la réserve de 

graines et fruits demeure suffisante pour exprimer une certaine 

richesse spécifique certaines années. La présence de tourbe, malgré 

sa mise en culture saisonnière, est favorable à cette expression. Elle 

subit périodiquement les inondations de l’oued El Kébir-Est, ce qui 

explique son caractère mixte : formations à Isoetes velata (dunes) et 

à Scirpus maritimus (plaines sublittorales). 

* Hrib Nord (16) est une ancienne carrière de sable, et cette 

mare est périodiquement connectée à la précédente, située en aval. 

Sa texture sableuse est du même ordre que la précédente (92 %) 

avec un pH très acide (4.20). Souvent asséchée et cultivée, cette 

mare exprime certaines années une très grande biodiversité, y 

compris certaines Orchidées. Aux espèces, citées dans l’analyse, 

s’ajoutent quelques espèces banales, comme Euphorbia helioscopia, 

Linaria pinnifolia ou Senecio leucanthemifolius. 

* Tamaris (17) appartenait à un ensemble assez proche de 

celui du groupement à Isoetes, témoin le nombre d’individus 

d’Isoetes velata présents à proximité ; cet ensemble, très marqué par 

la mise en culture, était naguère occupé par des bosquets de Quercus 

suber. La fraction sableuse est élevée (84 %). Comme la plupart 

de ces hydrosytèmes dunaires, elle est riche en matière organique, 

donc avec une dominante de l’azote (C/N = 5.89). La frondaison 

d’un individu de Quercus suber centenaire recouvre partiellement 

ce site. Imperata cylindrica, Arundo donax, Hypericum humifusum, 

Aster squamatus, Inula graveolens, Bellis annua, Alliaria officinalis et 

Chenopodium album sont fréquents autour de cette mare. 

* Carrière (18) a été utilisée pour l’extraction du sable, et 

elle permet l’abreuvement des troupeaux et l’irrigation des terres 

cultivées environnantes. Elle est parfois alimentée par une source, 

issue de la nappe dunaire. Sableuse à 93 %, son pH est acide (5.95). 

Elle est la seule où ont été trouvés Damasonium alisma et, avec les 

Salines, Zanichellia palustris subsp. pedunculata (sous-espèce à 

confirmer, étant donné les conditions écologiques très différentes 

des Salines). 

* Mafragh (19) fait partie d’un complexe de mares, favorisé 

par l’altitude faible (à peine 2 m au-dessus du niveau de la mer). 

Au nord-ouest du marais du Mkrada et à proximité de l’exutoire 

de 3 oueds (oued Mafragh), un certain nombre de diaspores sont 

véhiculées par les inondations hivernales avec des pénétrations 

maritimes possibles ce qui favorise une relative richesse spécifique. 

Sa texture est dominée par l’argile à 93 % avec un pH alcalin de 8.35. 

Entièrement entourée naguère par la cocciféraie, ce sont désormais 

les Joncs qui dominent dans cette plaine basse aux nombreuses 

petites dépressions inondables. Le développement des Characeae 

s’avère remarquable et ces taxons peuvent occuper la totalité de la 

strate végétale immergée. 

* Sangliers (20) est une dépression située en interface entre les 

dunes et la plaine sublittorale. Sa texture et son pH sont presque 

identiques à ceux de la Mafragh. Cependant, à la différence de la 

précédente, elle est largement entourée de végétation, y compris 

arbustive, puisqu’une cocciféraie la domine sur trois côtés. Elle 

a subi plusieurs incendies, probablement pour permettre la 

repousse de graminées palatables. Les Characeae s’y développent 

régulièrement, tapissant le fond du site. 

Ces deux stations ont un statut particulier par rapport aux autres 

mares dunaires. Elles appartiennent à la partie basse de la plaine 

de la Mafragh, proche de la mer, à une altitude très faible (entre 2 

et 3 m au-dessus du niveau de la mer) et elles présentent donc un 

pH alcalin. 

4. Collines gréseuses 

* Fedjouj (11) est une mare artificielle, utilisée dans les années 

1990 comme carrière de tout-venant. Localisée au creux de collines 

gréseuses (aval du djebel Koursi), sa texture est sablo-argileuse 

(57 % de sable et 20 % d’argile). Le cortège floristique environnant 

est celui des subéraies, accompagné d’Inula viscosa et I. graveolens. 

En l’espace de 10 années, cette mare, très pauvre en espèces, s’est 

considérablement enrichie, y compris en espèces rares, comme 

Eleocharis multicaulis ou Fuirena pubescens et même en Characeae. 


Faits de conservation en Méditerranée 

Mediterranean Conservation News 

Assessing favourable and unfavourable areas 

for Bonelli’s Eagle in Spain 

and its conservation implications 

To know and understand where and why species occur is crucial 

to manage biodiversity, and a necessary precursor for schemes 

to mitigate population decline. The development of relatively 

simple species distribution models is particularly interesting in the 

case of endangered species. Bonelli’s Eagle, Hieraaetus fasciatus 

(Vieillot, 1822), is a widespread raptor whose Western Palaearctic 

populations are distributed mainly in the Mediterranean area. 

In recent decades this species has suffered a severe population 

decline, and has been listed as an endangered European species. 

In Spain, which with 650-713 breeding pairs supports about 70 % 

of the European population, local extinction rates ranging from 

32.1 % to 48.6 % have been reported, and the species has recently 

changed its status from Vulnerable to Endangered (IUCN categories). 

The main reported causes of the decline are direct persecution, 

and electrocution by and collision with electric power 

lines. At present, European (Council Directive 79/409/EEC) 

and Spanish (Real Decreto 439/1990) legislation includes it as a 

priority target species for special conservation measures. 

Investigation on the factors that affect Bonelli’s eagle has been 

mostly based on local-scale ecological studies. However, Bonelli’s 

eagle populations are not only affected by local habitat characteristics, 

but also by environmental, and human-related processes 

that act on larger geographical scales. Consequently, broad-scale 

distribution models may help conservation programs to attain 

more satisfactory results, as the factors that affect the populations 

on a larger scale are taken into account. 

We predicted the potential distribution of Bonelli’s Eagle in 

Spain, performing a stepwise logistic regression, using 29 independent 

variables related to spatial situation, topography, climate, 

lithology, and human activity, and the presence/absence data on 

the UTM 10x10-km squares as geographic units. We also classified 

each UTM 10x10-km square into three categories (favourable, 

unfavourable and of intermediate favourability) depending 

on the favourability values yielded by the model. 

Mean slope, temperature of July and precipitation were found 

to correctly predict more than 76 % of Bonelli’s Eagle presences 

and absences in Spain. Although the species presence is more 

likely as the distance to highways and to big cities increases, the 

human activity plays a secondary role in Bonelli’s Eagle distribution 

at the considered scale. The fragmented spatial structure of 

the favourable areas in our distribution model suggests the exis- 

tence of a metapopulation dynamics, superimposed to the sourcesink 

dynamics implicated in the distribution of the species. 

From January 1997 to June 2006 Bonelli’s Eagle has benefited 

from LIFE projects an amount of nearly 8 millions (http: 

//europa.eu.int/comm/environment/life/home.htm), co-financed 

by the European Union (68.7 % of the budget) and the Spanish 

Government. These projects are generally focused on those areas 

where a huge decline has been recorded, in the limit of its range, 

whereas southern and south-eastern populations remain apparently 

stable. We propose to pay attention also to southern favourable 

areas, where the auspicious status of the species could be 

deceptive, since an increase in the percentage of pairs with at least 

one non-adult has been detected. The conditions of southern and 

more favourable areas could also explain the decline of northern 

populations if they fail in exporting enough emigrants. So, those 

actions developed in unfavourable areas, favoured through LIFE 

projects, should be complemented with actions in favourable 

areas, which might favour population persistence in unfavourable 

areas through dispersal processes. 

A. ROMÁN MUÑOZ 

DEPARTAMENTO DE BIOLOGÍA ANIMAL, FACULTAD DE CIENCIAS, 

UNIVERSIDAD DE MÁLAGA, E-29071 MÁLAGA, SPAIN. 

E-MAIL: roman@uma.es 

For more information, read: 

Muñoz, A. R.; Real, R.; Barbosa, A. M. & Vargas, J. M., 2005. Modelling 

the distribution of Bonelli’s Eagle in Spain: Implications for conservation 

planning. Diversity and Distribution, vol. 11: 477-486. 

What is required for pond turtles to survive? 

More than just a pond 

The European pond turtle, Emys orbicularis, is a species having 

strong conservation concern. It has declined in all the European 

countries were it lives: actually, in many areas it is very rare. The 

European Union founds actions for the protection of this species 

and for the management of habitats. It is clear that prior to start a 

management plan, the causes of the decline of a species should be 

addressed, to optimise the available founding and to increase the 

possibility of success. To date, most of actions for the protection 

of the “pond” turtle Emys orbicularis have focused on the conservation 

of wetlands. However, in the last years several studies, 

mainly in North America, outlined the importance of terrestrial 

101 


◆ FAITS DE CONSERVATION EN MÉDITERRANÉE / MEDITERRANEAN CONSERVATION NEWS 

102 

habitat for freshwater turtles, and suggested that a more broad 

scale approach should be more appropriate for their protection. 

Therefore, we evaluated the importance of terrestrial habitat for 

the European pond turtle, by measuring turtle presence and 

abundance in a large number of wetlands in the Po River Delta, 

a large protected wetland system in Northern Italy. 

Despite some of the wetland features were important for 

this turtle (for example, E. orbicularis prefers permanent, large 

wetlands), we did not find a relationship between water features 

and turtle distribution: we frequently observed E. orbicularis also 

in wetlands having very eutrophic water. Surprisingly, we found 

that the composition of landscape surrounding a wetland was 

apparently more important than the wetland features: turtles were 

more frequent and more abundant in wetlands surrounded by 

woodlands. This result suggests that the management of habitats 

surrounding the wetlands is very important for this species; however, 

too often the management of semi aquatic species focused 

only in the aquatic component of their habitat. 

It should not be forgotten that vital activities for these animals, 

such as nesting, can occur far from the wetlands: for example, 

females has been observed laying their eggs thousand of meters 

far from their pond. Emys orbicularis is a long-lived animal. It 

can survive for decades in areas having the minimal requirements 

for the survival of adults. Sometimes, in the human dominated 

lowlands of Mediterranean Europe, it is possible to observe 

E. orbicularis adults in ponds completely surrounded by crops 

and human settlements. However, if the surrounding habitat is 

not suitable for reproduction and it does not allow among-pond 

movements, the fate of these isolated individuals is doomed. A 

large scale, landscape ecology approach is required for the protection 

of this endangered turtle. Unfortunately, the landscape of 

Mediterranean Europe, where E. orbicularis was once abundant, is 

now dominated by human activities: the protection of this species 

is therefore a challenge for conservationists and land managers. 

GENTILE FRANCESCO FICETOLA 

DIPARTIMENTO DI BIOLOGIA, UNIVERSITÀ DEGLI STUDI DI MILANO. V. 

CELORIA 26, 20133 MILANO ITALY. 

E-MAIL: francesco.ficetola@unimi.it 


Ficetola, G.F., Padoa-Schioppa, E., Monti, A., Massa, R., De Bernardi, F. 

& L. Bottoni 2004. The importance of aquatic and terrestrial habitat for 

the European pond turtle (Emys orbicularis): implications for conservation 

planning and management. Can. J. Zool. 82, 1704-1712. 

Butterflies and plants highlight traditional 

hay meadows as one of the most valuable and endangered 

habitats on the Spanish Mediterranean coast 

Situated in north-east of Catalonia, the Aiguamolls de l’Empordà 

Natural Park is one of the most outstanding wetland areas 

in the Iberian Peninsula and, indeed, in the whole Mediterranean 

basin. Traditionally highly regarded for its birdwatching potential, 

it was declared a Natural Park in 1983 and then as a Ramsar Site 

as a Wetland of International Importance. The initial impression 

gained of this protected area is one of a vast marshland; however, 

a closer examination reveals a mosaic of different habitats all 

Fig. 1. 3 males of the butterfly 

Plebejus argus, one of the most 

typical species inhabiting the 

« closes » and declining (photo 

Marta Miralles). 


FAITS DE CONSERVATION EN MÉDITERRANÉE / MEDITERRANEAN CONSERVATION NEWS ◆ 

Fig. 2. A typical « closa » in 

the study area (photo Marta 

Miralles). 

103 

associated with a myriad of differing environmental conditions. 

Recently, botanists have undertaken systematic surveys aimed at 

identifying which habitats are the most interesting in terms of 

their conservation value. Their results have revealed the exceptional 

importance of the so-called closes, the meadows enclosed 

by tree-lined drainage canals that flood in winter and are cut for 

hay once or twice during the year. These traditional hay meadows 

harbour the most diverse vegetation and also the region’s rarest 

taxa. However, the closes are at risk because they are no longer 

profitable and are being converted into arable fields or being 

abandoned and invaded by shrubs and scrub. The frequent lack 

of agreement in cross-taxonomic patterns of species richness and 

rarity at the fine geographical scale considered here encouraged us 

to assess the value of the closes through an independent data set. 

In particular, we used data from butterflies, an excellent indicator 

taxa whose populations have been monitored over the last 15 

years as part of the Catalan Butterfly Monitoring Scheme. Our 

analyses showed a strong link between the various habitat types 

and the composition of butterfly assemblages and, once again, 

pointed to the closes as the most interesting habitat in the Natural 

Park: they hold more butterflies and a tendency to boast rarer 

species. The coincidence in species richness and rarity patterns 

across two groups of taxa encompassing two different trophic 

levels, together with the incredibe speed at which the closes are 

vanishing, highlight the necessity of placing these traditional 

hay meadows firmly at the top of the agenda of Spanish and 

European conservation bodies. It has been estimated that 80 % of 

the closes have been lost over the last 50 years and – even more 

incredibly – that at least 60 % have disappeared since the area 

became legally protected in 1983. In fact, they are paradigmatic 

of a common problem faced by most Mediterranean protected 

areas: a lack of effective financial support that is leading eventually 

to the loss of unprofitable traditionally man-managed habitats. We 

believe that the future of the closes and other grassland habitats 

supporting an exceptionally rich wildlife in the Mediterranean 

basin will ultimately depend on the existence of agri-environmental 

schemes sponsored by the governments of individual 

countries and/or by projects developed at a broader level within 

the European Union. 


CONSTANTÍ STEFANESCU 

BUTTERFLY MONITORING SCHEME 

MUSEU DE GRANOLLERS CIÈNCIES NATURALS 

FRANCESC MACIÀ, 51, 08400 GRANOLLERS, SPAIN 

E-MAIL: canliro@teleline.es 

Stefanescu, C., Peñuelas, J. & Filella, I. 2005. Butterflies highlight the conservation 

value of hay meadows highly threatened by land-use changes in a 

protected Mediterranean area. Biol. Conserv., 126: 234-246. 

Prey detectability, more than prey density, 

affects diet composition of Bonelli´s eagle 

Hieraaetus fasciatus: conservation implications 

Food availability in one of the most important factors influencing 

the quality of raptor habitats, which is determined not only 

by prey density, but also by the accessibility to prey by predators. 


◆ FAITS DE CONSERVATION EN MÉDITERRANÉE / MEDITERRANEAN CONSERVATION NEWS 

104 

In the diet of raptors the presence of prey species is influenced 

by several factors, such as their abundance and the ground-level 

vegetation in territories. 

We analysed this situation for the Bonelli´s eagle (Hieraaetus 

fasciatus) in south-eastern Spain, a largely mountainous region with 

typical Mediterranean vegetation and a healthy population of the 

Bonelli´s eagle. First, we performed a dietary analysis of the whole 

population, which indicated that Oryctolagus cuniculus (29.1 %) and 

Alectoris rufa (28.8 %) were the main prey of the eagle, followed by 

Columba palumbus (13.3 %) and Columba livia (8.0 %). Second, 

we tested for the minimum number of prey items for the reliability 

of results, finding between 15 and 30 prey items, depending on the 

pair. Third, we checked for differences in prey frequency and productivity 

among pairs, discovering strong differences in both traits, 

but we failed to find any relationship between the frequency of the 

main preys in the diet and productivity (mean productivity of the 

pairs, 1.4 nestling/year). Fourth, we found that only the percentage 

of European wild rabbit in diet was correlated with its abundance in 

territories (r = 0.81; P = 0.01) although, the percentage of rabbits 

in the diet was better correlated with the amount of open land in 

the territories (r = 0.87; P = 0.005). Thus, the percentage of open 

land within the territories was the single variable, selected by a 

multiple-regression analysis, that explained the frequency of rabbits 

in the raptor’s diet (F 2,5 = 14.64, R 2 = 0.85; P < 0.008); the same 

was true for the stepwise forward (F 2,5 = 14.64, R 2 = 0.85; P < 

0.008) and backward multiple-regression analysis (F 1,6 = 16.67, R 2 

= 0.74; P < 0.006). These results suggest that accessibility of this 

prey type would be more important than absolute abundance for 

the Bonelli’s eagle. This would be important for the management 

of some Bonelli´s eagle populations, since prey accessibility (depending 

on ground-level vegetation) is determinant for the capture of 

rabbits, its main prey species in Spain. As expected for a raptor that 

relies on few and prominent prey species, the presence of rabbits in 

the diet influenced other prey frequencies. We found negative and 

significant relationships between the presence of rabbits versus red 

legged partridge (r = -0.81; P = 0.01) and versus wood pigeons 

(r = -0.89; P = 0.003) in diet. Birds as prey appear to complement 

the diet when open-land scarcity in the territories implies low rabbit 

detectability and consumption. 

The results for southern and healthy populations (uncommon 

circumstance in this species), indicate the importance of alternative 

preys (partridges and pigeons) in some territories with low open-land 

surface area; for northern and more threatened populations, these 

results suggest the importance of adequate habitat management. 

Conservation measures proposed concerning the increase of prey 

availability in areas with declining populations should consider both 

the absolute prey density and prey detectability, favouring a vegetation 

structure adequate for prey detection and hunting success of the 

eagle and avoiding extensive reforestation in those territories. 

Scientifically-based management solves conflict 

between European storm petrels and their facultative 

predator at minimum cost 

Large gulls are commnonly perceived as pest species by environmental 

managers owing to their facultative predatory habits 

on smaller bird species (e.g. Finney et al., 2003). Once predation 

is detected, conservation actions typically involve indicriminated 

gull culling in the absence of any proof of negative influence of 

predation on the growth rate of prey populations. Here we report 

on a successful management plan based on the results of an ad hoc 

research program carried out in a small western Mediterranean 

island (i.e. the island of Benidorm) holding both a storm petrel 

and a yellow-legged gull colony of about 500 breeding pairs 

each. 

First, based on the long-term monitoring of the storm petrel 

colony, the effects of predation on the population dynamics of the 

tubenose birds were explored through capture-recapture analysis 

(Oro et al., 2005). Before management actions, 122 and 124 birds 

were killed by gulls in 2002 and 2003 respectively. Results showed 

that as much as 10 % of the birds killed by gulls were breeding 

adults and that predation was indeed an additive cause of mortality 

of adult birds. This was the case starting since the year of 

operation of a powerful public lighting installed along the nearby 

touristic coastline that made it easier for gulls to predate on petrels 

at night. Hence, management actions had to be applied urgently, 

considering that reductions in adult survival of long-lived birds 

have the largest influence on their population growth rates (e.g. 

Saether et al., 1996). 

DIEGO ONTIVEROS, JUAN M. PLEGUEZUELOS AND JESÚS CARO 

ADDRESS: DEPARTAMENTO DE BIOLOGÍA ANIMAL 

Y ECOLOGÍA, FACULTAD DE CIENCIAS, 

UNIVERSIDAD DE GRANADA, GRANADA E-18071. 

E-MAIL: dontive@ugr.es 

Fig. 1. A breeding storm petrel. 


FAITS DE CONSERVATION EN MÉDITERRANÉE / MEDITERRANEAN CONSERVATION NEWS ◆ 

Secondly, gulls were monitored as to individuals containing 

storm petrels in their diet among the whole local population. 

Results showed unequivocally that only a small number of gulls 

were responsible of most predation, more especifically those whose 

nests were located close to the entrance of the two caves where 

most storm petrels breed (Mínguez et al., 2005). Management 

actions were therefore addressed to removing only those predator 

pairs, leaving the bulk of gulls untouched. By means of 

nest traps 11 gulls were removed in 2004 and 18 in 2005. The 

removal of these individuals led to a surprising reduction in the 

number of gull pellets found inside storm petrel colonies and 

along a transect between them in 2004 and 2005 (51 and 75 % 

respectively), compared to the situation before gull removal (Sanz 

et al., in prep.). A number of years iares still necessary to quantify 

more precisely the decrease in predation on adult survival using 

capture-recapture modelling, but results show unquestionably the 

success of the control. 

Management actions should probably be implemented on 

an annual basis since the niche of gulls facultatively predating 

on storm-petrels seems to be readily occupied by other gulls, 

provided that due changes in public lighting are not performed 

in the future. Although the implementation of these tasks is both 

time and resource consuming, its costs are neglegible compared 

to the traditional solution of performing massive gull cullings 

when faced with similar conflicts, not to mention social costs. 

This small-scale case study clearly illustrates the tremendous 

potential of scientifically-based wildlife management in terms 

of resource optimization and successful achievement of desired 

conservation goals. 

Literature cited: 

Finney S.K., Harris M.P., Keller L.F., Elston D.A., Monaghan P. & 

Wanless S., 2003. Reducing the density of breeding gulls influences the 

pattern of recruitment of immature Atlantic puffins Fratercula arctica 

to a breeding colony. Journal of Applied Ecology, 40: 545-552 

Mínguez, E., Sanz, A., de León, A. & Junza, M., 2005. No todas las gaviotas 

comen paíños. Quercus 228: 81. 

Oro, D., De León, A., Minguez, E., & Furness, R.W., 2005. Estimating 

predation on breeding European Storm-petrels by yellow-legged gulls. 

Journal of Zoology 265: 421-429. 

Saether B.E., Ringsby T.H. & Roskaft E., 1996. Life history variation, 

population processes and priorities in species conservation: Towards 

a reunion of research paradigms. Oikos 77: 217-226 

ALEX DANI, ANA SANZ 1 , EDUARDO MÍNGUEZ 1 , LLANOS DE LEÓN 2 , 

ALEJANDRO MARTÍNEZ-ABRAÍN 3 , BLANCA SARZO 4 , 

ELENA VILLUENDAS 4 , JOSÉ SANTAMARÍA 4 AND DANIEL ORO 3 

1. UNIVERSIDAD MIGUEL HERNÁNDEZ 

2. UNIVERSITY OF GLASGOW 

3. IMEDEA (CSIC-UIB) 

4. CONSELLERIA DE TERRITORI I HABITAGE 

CORRESPONDING AUTHOR : 

DR. ALEJANDRO MARTÍNEZ ABRAÍN, POPULATION ECOLOGY GROUP 

IMEDEA (CSIC-UIB), AVDA. DE LOS PINARES 106, 

46012 EL SALER, VALENCIA (SPAIN) 

E-MAIL: a.abrain@uib.es 

http://www.imedea.uib.es/natural/goi/seabirds 

105 


Analyses d’ouvrages 

Atlas of climatic diagrams 

for the isoclimatic mediterranean zones 

H.N. Le Houérou 

220 p., (2004) ISBN : 2-9523965-0-7 

Prix 35,5 euros + frais d’envoi. Disponible chez l’auteur, 

327 rue A.-L. de Jussieu, F-34090, Montpellier. 

hn.le-houerou@club-internet.fr 

Cet atlas comprend 1 560 diagrammes de 1 320 stations climatiques 

de 60 pays ayant totalement ou partiellement des climats 

méditerranéens, c’est-à-dire à pluies d’hiver et sècheresse estivale, 

tel le bassin méditerranéen. Ces types de climats sont répartis sur 

tous les continents, sauf l’Antarctique, entre les latitudes de 25 et 

46°N et S. Ils représentent une superficie de quelque 15 millions 

de km 2 , c’est-à-dire un peu plus de 11 % des masses continentales 

de la planète. L’atlas comprend 220 pages, dont 170 sont des 

planches de graphiques ombrothermiques et ombrodiapnéiques 

(ETP). Les graphiques sont groupés par ordre alphabétique de 

pays et de station, ce qui facilite la consultation. L’atlas comprend 

aussi une classification bioclimatique fondée sur deux critères 

principaux : un indice d’aridité (P/ETo) : le quotient de la pluviosité 

moyenne annuelle par l’évapotranspiration potentielle de 

référence. Le second critère (m) est la moyenne des températures 

minimales journalières du mois le plus froid. Cette classification 

est donnée à la fois dans un graphique orthogonal et un tableau ; 

elle comprend 155 stations en provenance de 50 pays. Des explications 

sur la construction des diagrammes, leur interprétation, la 

liste et les superficies des zones à climat méditerranéen occupent 

19 pages, la bibliographie comprend 136 titres, et un index de 28 

pages complète le volume. 

Les diagrammes ombrothermiques montrent la marche mensuelle 

des précipitations, de la température et de l’évapotranspiration 

potentielle ; ils permettent ainsi d’évaluer la longueur et 

l’intensité des saisons sèche et pluvieuse, la durée et la sévérité des 

périodes de repos hivernal des plantes du aux basses températures. 

Il existe plusieurs types de diagrammes. Le modèle le plus simple 

est celui proposé par Bagnouls et Gaussen en 1953, popularisé 

par le Klimadiagramm Weltatlas de Walter et Lieth, publié en 1960, 

mais épuisé depuis des décennies. Ce modèle a été utilisé par des 

milliers de chercheurs et techniciens de diverses disciplines dans 

diverses parties du monde depuis 50 ans, pour la classification 

des climats. D’autres modèles utilisent en outre l’évapotranspiration 

potentielle de référence, ce qui permet d’établir des bilans 

hydriques climatiques. Ces derniers types de diagrammes sont au 

nombre de 350 ; ils sont originaux ou résultent des publications 

antérieures de l’auteur. Beaucoup ont été construits à partir de 

la base de données CLINO, publiée par l’OMM en 1996. Des 

comparaisons zonales et continentales sont proposées pour des 

zones ayant des diagrammes particulièrement semblables tels que 

ceux de la Californie et du Chili, de la Californie et du bassin 

méditerranéen ou du Great Basin, de la région irano-touranienne, 

de l’Asie moyenne et de la Patagonie argentine. Des comparaisons 

sont aussi offertes avec d’autres types de classifications, telles celles 

basées sur l’altitude, en honneur dans certains pays du bassin 

méditerranéen. Des zones marginalement méditerranéennes, 

telles le sud de la Crimée et certaines zones des Balkans, sont 

incluses dans l’atlas. De même, certaines vallées à climat méditerranéen 

de l’Hindoukouch et du SW de l’Himalaya ainsi que la 

limite entre les climats méditerranéens et tropicaux dans le souscontinent 

indien. Une des originalités de cet atlas est la place faite 

à la région irano-touranienne qui s’étend de la rive orientale de la 

Méditerranée jusqu’à la frontière occidentale de la Chine sur le 

méridien 76°E, où le régime à précipitations hivernales de l’Asie 

moyenne cède brusquement la place à un régime à pluies d’été 

(régime de mousson) qui caractérise le Xinjiang et la Mongolie. 

Cet atlas peut être considéré comme un outil de référence 

utile pour les géographes, les biogéographes, les écologues, les 

forestiers, les pastoralistes et les agronomes, c’est-à-dire tous ceux 

pour qui la caractérisation bioclimatique est une nécessité. 

The isoclimatic mediterranean biomes: 

Bioclimatology, diversity 

and phytogeography 

H.N. Le Houérou 

765 p., (2005) ISBN : 2-9523965-1-5 

Prix 65 euro + frais d’envoi. Disponible chez l’auteur, 

327 rue A.-L. de Jussieu, F-34090, Montpellier. 

hn.le-houerou@club-internet.fr 

Il s’agit d’un ouvrage de 765 pages en deux volumes. Le 

premier volume (365 p.) contient le texte, une liste de références 

de 1 600 titres et les remerciements. Le second volume (400 p.) 

contient les annexes, dont 103 tableaux hors-texte, 118 figures, 

un glossaire de 1 900 termes et un index de 650 entrées. 

Cet ouvrage fait partie d’une trilogie dont les deux autres 

termes sont publiés par ailleurs : 

— Atlas des diagrammes climatiques des stations de la zone isoclimatique 

méditerranéenne, 220 p., 1 560 graphiques de 60 pays. 

107 


◆ ANALYSES D’OUVRAGES 

108 

— Atlas de la répartition de 250 espèces clés dans le bassin méditerranéen. 

150 p., 300 cartes et schémas. Disponible fin 2005 au 

CIHEAM-IAMZ. Saragosse. 

Les biomes se définissent comme les grandes zones continentales 

de milieux naturels : la Toundra, la Taïga, la Forêt tempérée, 

la Prairie, la Steppe, la Forêt tropicale, la Savane, la Forêt pluviale 

équatoriale et les Déserts. Ces biomes se caractérisent par un 

climat défini, une flore, une végétation, une faune, des sols, une 

géomorphologie, des cultures, un mode d’habitat et d’exploitation 

des terres, etc. La zone isoclimatique méditerranéenne se 

définit comme l’ensemble des régions recevant des précipitations 

hivernales et subissant une sécheresse plus ou moins prononcée 

l’été. Elles possèdent une végétation sclérophylle dans les zones 

semi-arides à hyper-humides, steppique dans les zones arides, et 

contractée dans les zones hyper-arides ou désertiques. Les difficultés 

surgissent lorsqu’il faut décider de l’abondance relative des 

précipitations hivernales et de l’absence partielle (ou totale) des 

précipitations estivales pour qu’une zone déterminée puisse être 

qualifiée de méditerranéenne. Pour ce faire, l’auteur définit deux 

indices objectifs de « méditerranéité » IM 1 et IM 2 . IM 1 est le 

rapport entre les précipitations du trimestre hivernal et celles du 

trimestre estival (IM 1 = PTH / PTE). IM 2 et le quotient des 

précipitations du semestre hivernal (à jours courts) à celles du 

semestre estival (à jours longs) (IM 2 = PSH / PSE). On peut 

aussi invoquer, ce qui revient au même, le % des précipitations du 

trimestre hivernal dans le total annuel : par définition, les pluies 

du trimestre hivernal sont supérieures à 25 % du total annuel 

et les pluies semestrielles d’hiver à plus de 50 % des chutes 

annuelles. Par approximations successives et en se fondant sur la 

nature et la répartition de la flore, de la végétation, de la faune, 

des systèmes d’élevage et des cultures, l’auteur est arrivé à la 

conclusion suivante : pour qu’une zone puisse être qualifiée de 

méditerranéenne il faut que IM 1 > 2,0 ; il peut atteindre l’infini 

lorsque le total des pluies du trimestre estival est nul, comme c’est 

souvent le cas de basses terres de la Méditerrannée orientale et 

du Proche-Orient, ainsi qualifiées d’« hyper-méditerranéennes ». 

L’IM 2 doit dépasser 1,5. Le critère de température hivernale, 

utilisé par certains auteurs, n’entre pas en ligne de compte à ce 

niveau, mais plus en avant dans les critères de classification. Les 

cas litigieux sont résolus par l’examen détaillé de la végétation 

naturelle et des cultures, mais il reste, bien entendu, des zones 

de transition appelées sub-méditerranéennes (exemples : les 

Causses, les Alpes maritimes, les Apennins, la partie orientale de 

la chaîne des Pyrénées). Les spécialistes peuvent ainsi constater 

que l’auteur s’est largement inspiré à la fois des concepts de ses 

maîtres Emberger et Gaussen, en les adaptant aux connaissances 

modernes, sur l’ETP, par exemple. Ainsi définies, les régions 

méditerranéennes couvrent près de 15 millions de km 2 (30 fois la 

France) et représentent près de 12 % des terres émergées dans 60 

pays ou États partiellement ou totalement méditerranéens (16 % 

de la superficie nationale en France). 

Exemples : les 20 pays adjacents à la Méditerranée et le Portugal 

(avec les îles Insulo-Atlantiques orientales ou macaronésiennes), 

l’Asie moyenne (1/3 sud du Kazakhstan, Kirghizistan, 

Ouzbekistan, Tadjikistan, Turkménistan), le Proche-Orient 

(Turquie, Syrie, Israël, Palestine, Jordanie), Moyen-Orient (Irak, 

Iran, Afghanistan, Pakistan à l’ouest de l’Indus), les 2/3 NE de la 

Péninsule arabique, toute l’Afrique du Nord, la partie sud-ouest de 

l’Afrique du Sud, le long de l’Atlantique, les îles Canaries, Madère 

et les Açores, la Californie, une grande partie de l’Orégon, de l’état 

de Washington, de l’Idaho, du Nevada, la moitié ouest de l’Utah 

(tout l’ouest du Great Basin), l’extrémité sud-ouest de l’Arizona, 

la basse Californie du Nord (France), le coin SO de la Colombie 

Britannique (Vancouver), le Chili central entre les 25° et 35° de 

latitude sud, une grande partie de la Patagonie argentine et les 

piedmonts orientaux des Andes entre les 30° et 45° parallèles sud. 

Presque toute l’Australie du Sud, 1/4 sud-ouest des Nouvelles Galles 

du Sud et la moitié occidentale de Victoria et 1/3 SO de l’Australie 

de l’Ouest. On a ainsi montré la grande similitude bioclimatique 

entre la partie ouest du Great Basin, la région Aralo-Caspienne et 

la Patagonie, confirmée, si besoin était, par des introductions de 

plantes réussies et réciproques et par l’invasion d’espèces iranotouraniennes 

dans le Great Basin. Divers aspects des climats 

méditerranéens sont étudiés en détail montant et saisonnalité des 

précipitations annuelles, évolution à long terme des précipitations 

annuelles, variabilité annuelle (inversement proportionnelle à la 

hauteur, mais variable dans une large proportion d’une région 

à l’autre). Les climats méditerranéens présentent deux critères 

essentiels pour la vie et la répartition des plantes et des animaux : 

le bilan entre l’offre et la demande d’eau (indice d’aridité) et le 

stress thermique représenté par le froid hivernal (ou son absence). 

Le premier se mesure par le rapport entre les précipitations et 

l’évapotranspiration potentielle (évaluée par lysimètre ou calculée 

au moyen de l’équation de Penman, (P/ETPp) et le second par la 

moyenne des températures minimales journalières du mois le plus 

froid (janvier dans l’hémisphère Nord, juillet dans l’hémisphère 

Sud). Les précipitations n’offrent aucune tendance évolutive à long 

terme au cours des 170 ans pour lesquels des mesures existent. 

Mais il existe des tendances positives et négatives à moyen terme 

(25-50 ans). La température planétaire à long terme a augmenté 

d’environ 0,5 °C en 100 ans ; elle est probablement de moins de la 

moitié de ce chiffre pour les latitudes méditerranéennes (25°-45° N 

et S), ce qui correspond à l’augmentation de température attribuée 

à l’urbanisation, mais elle est beaucoup plus élevée au-delà des 

latitudes de 45° (de 1 à 1,5 °C entre 50° et 60° de latitude N et S). 

De fait, aucune évolution à long terme de la température n’a encore 

pu être mise en évidence sous les latitudes méditerranéennes quel 

que soit le continent considéré, ce qui n’implique pas qu’elle ne le 

soient dans un avenir prévisible. 

L’indice d’aridité (P/ETPp) permet de distinguer 7 zones 

méditerranéennes principales en fonction de l’aridité, selon la 

terminologie d’Emberger, reprise dans la carte mondiale des 

zones arides de l’Unesco : 


ANALYSES D’OUVRAGES ◆ 

— hyperhumide (P > ETPp), 

— humide (ETPp > P > 0,70 ETPp) , 

— sub-humide (0,70 ETPp > P > 0,45 ETPp), 

— semi-aride (0,45 ETPp > P > 0,28 ETPp), 

— aride (0,28 ETPp > P > 0,07 ETPp), 

— hyper-aride (0, 07 ETPp > P > 0,03 ETPp) 

—et érémitique (0,03 ETPp > P), 

Les limites entre ces zones sont déterminées empiriquement 

par la répartition de la végétation naturelle, des cultures et des 

systèmes de production agricole et d’élevage et la géomorphologie, 

comme le préconisait Emberger. Les déserts méditerranéens 

se caractérisent par le fait que les rares pluies qui y tombent surviennent 

toujours en hiver, contrairement aux déserts tropicaux 

(à pluies d’été) et tempérés (à régime pluviométrique saisonnier 

équilibré). 

L’indice de stress thermique hivernal est la moyenne des 

températures minimales journalières du mois le plus froid (janvier 

dans l’hémisphère Nord, juillet dans l’hémisphère Sud). Ce 

paramètre est étroitement corrélé avec le nombre annuel de jours 

de gel, donnée plus difficile à rencontrer et souvent indisponible. 

Il permet de distinguer 8 tranches de 2 °C depuis l’extrêmement 

chaud (m > +9 °C) jusqu’à l’extrêmement froid (m < -5 °C). Les 

raisons du choix de ces critères sont données en détail ; on ne peut 

s’y étendre ici. Disons cependant que m = -5 °C correspond à la 

limite supérieure des arbres dans les hautes montagnes méditerranéennes 

et que m > +9 °C correspond à l’absence de gel sous abri 

et à la dominance concomitante d’espèces végétales et de cultures 

d’affinité tropicale lorsque l’eau est disponible. Dans le premier 

cas, la température moyenne mensuelle du mois le plus froid est 

d’environ 0 °C, et 15° C dans le second, mais cette relation est peu 

fiable. Ces tranches thermiques sont empiriquement justifiées par 

la présence ou l’absence de certaines espèces spontanées, de cultures 

et d’animaux. Nous avons ainsi une matrice orthogonale de 

(7 x 8 ) 56 bioclimats méditerranéens principaux. Des variantes et 

nuances peuvent être ajoutées en faisant intervenir la saisonnalité 

des précipitations, leur variabilité, les précipitations occultes, la 

moyenne des températures maximales journalières du mois le plus 

froid, le nombre de jours de gel, les dates moyennes du premier et 

dernier gel annuel, etc. Il s’agit donc d’un système de classification 

« ouvert » qui peut s’adapter aux nécessités locales, car on ne 

peut tout prévoir ! Ces différentes catégories peuvent s’identifier 

visuellement et de manière quasi instantanée par l’examen des 

graphiques ombro-thermiques ou ombro-diapnéiques montrant 

la marche mensuelle des précipitations et des températures et des 

évapo-transpirations potentielles (cette relation, inventée il y a 170 

ans par l’agronome A. de Gasparin est utilisée partout depuis 

les années 1950, suite à la « ré-invention » du concept par les 

botanistes et phytogéographes toulousains Gaussen et Bagnouls). 

Dans ces graphiques les précipitations moyennes mensuelles et 

l’ETPp sont représentées sur une échelle double des températures 

moyennes (P = 2 t), ou encore (P = 0,35 ETPp). Ce dernier 

critère a été établi par Le Houérou et Popov dans une étude sur 

la bioclimatologie de l’Afrique, publiée par la FAO en 1981. Les 

deux critères coïncident dans 98 % des cas en zone isoclimatique 

méditerranéenne. À partir de ces critères on déduit que statistiquement 

ETPp (mm) ~ 0,19 °C à l’échelle journalière pour 

les zones où le vent est faible (< 5 m/s en moyenne) soit ETPp 

~ 70 T à l’échelle annuelle. Une température moyenne annuelle 

de 20 °C correspond donc, pour ces zones peu ventées, à une 

ETPp annuelle de 1 400 mm/ an. 

Dans les zones très ventées, comme le Sahara ou la Patagonie, 

ce rapport peut être beaucoup plus élevé (20 à 60 % supérieur, 

en raison de l’importance locale du terme aérodynamique de 

l’équation de Penman dans l’évapotranspiration globale de référence 

(ETo). Ainsi définies, les zones méditerranéennes possèdent 

une flore d’environ 75 000 espèces vasculaires (plantes à fleurs 

et fougères) dont 54 % sont endémiques, c’est-à-dire limitées 

aux territoires méditerranéens. Ce nombre d’espèces représente 

25 % de la flore terrestre sur 12 % du territoire de la planète. 

Pour fixer les idées, disons que la France possède environ 4 500 

espèces, dont 3 500 sont présentes dans la zone méditerranéenne 

(2 800 en Corse), l’Espagne l’Italie et la Grèce possèdent quelque 

5 500 espèces chacune, l’Afrique du Nord 7 200, la Turquie 

11 000, Flora palaestina 3 000, etc. Par ailleurs, on estime que les 

5 % plus érudits botanistes peuvent instantanément reconnaître 

et nommer de mémoire environ 5 000 espèces. 

Certaines régions méditerranéennes offrent une richesse 

considérable : la région du Cap s.l. 12 000 espèces, l’Australie de 

l’Ouest 10 000. Ces deux territoire présentent un taux d’endémicité 

de 80 %. Le bassin méditerranéen au sens large comprend 

25 000 dont 60 % d’endémiques et la région irano-touranienne 

17 000 espèces, dont 30 % d’endémiques. En richesse aréale 

spécifique le Cap arrive très largement en tête avec 600 espèces 

par 10 000 km 2 et le Fynbos (équivalent des garrigues en 

Afrique du Sud) avec 1 000 espèces par 10 000 km 2 (France 90). 

Il faut noter que les très riches flores du Cap et de l’Australie de 

l’Ouest correspondent à des terres oligotrophes, très pauvres sur 

le plan de la nutrition minérale, avec des pH tombant parfois à 

3,0. Ce fait entraîne des adaptations très originales des systèmes 

d’assimilation minérale. Des espèces endémiques y sont parfois 

cantonnées à des zones de quelques km 2 . 

L’ouvrage étudie ainsi de façon détaillée, outre les climats, 

les flores, les végétations, les cultures (y compris ornementales) 

et la présence d’espèces exotiques naturalisées et leur importance 

dans les flores locales (environ 10 % du nombre d’espèces 

régionales, en moyenne). Les succès et les échecs de transfert 

transcontinentaux d’espèces d’une zone méditerranéenne à 

l’autre sont examinés. Ces échanges ont donné lieu à des succès 

spectaculaires en partie dus à l’émigration de populations humaines 

méditerranéennes au Nouveau Monde et en Australie (Acacia australiens, 

Agropyron spp, Amandier, Artichaut, Asperge, Atriplex, 

Eucalyptus, Luzerne, Olivier, Vigne, etc.). On note aussi l’invasion 

d’espèces comme le brome des toits ou cheat grass (Bromus 

109 


◆ ANALYSES D’OUVRAGES 

110 

tectorum), devenu au siècle dernier une peste dans les parcours 

du NW des USA ainsi que d’autres espèces irano-touraniennes. 

On note aussi des échecs cuisants (Tamarugo, Jojoba, Maireana, 

Mulga). Ces échecs résultent en général d’un manque de prise 

en compte suffisant des spécificités bioclimatiques ou édaphiques 

des espèces concernées. 

La classification phytogéographique proposée prend en considération 

les critères climatiques, floristiques, végétationnels et 

agronomiques. Elle élève la zone phytogéographique méditerranéenne, 

jusqu’ici considérée comme une simple région, au rang 

supérieur de royaume. Nous avons ainsi : un empire holarctique, 

lequel inclut les royaumes euro-sibérien et méditerranéen, entre 

autres. Ce dernier comprend 7 régions : bassin méditerranéen, 

saharo-arabique, Asie moyenne (aralo-caspienne ou iranotouranienne), 

région du Cap, Californie et Great Basin, Chili- 

Argentine, Sud-Est et Sud-Ouest australiens. Cette classification 

est différente dans sa conception de celle de Takhtajan, mais reste 

compatible avec elle dans son résultat. 

Restoration Ecology: the new frontier 

J. Van Andel & J. Aronson (eds.) 

Blackwell Publishing, USA: 319 p. (2006) 

This book explores the interface between restoration ecology 

and ecological restoration. It aims at introducing to interactions 

between theory and practice. It challenges ecologist to explore the 

applicability of current theories and concepts, recognizing that 

these have not been developed with such applications in mind. 

The academic foundations of restoration ecology are revisited for 

this purpose, to pave the way towards a review of the causes of 

successes and failures and to identify the perspectives of ecological 

restoration in different ecosystem types. These are dealt with 

biome-by-biome and considered from the historical perspective 

of land use. The final section addresses problems of ecological 

restoration in a societal context, ecological restoration is meant 

to achieve sustainable, resilient and interconnected ecosystems 

and socioecological systems. This will result in newly emerging 

ecosystems. 

The originality of this book comparing to the previous 

published in restoration ecology is that a lot of chapters are 

case studies from Europe and not only from North-America. 

Nevertheless only one chapter is devoted to Mediterranean ecosystems 

(Restoration of Mediterranean woodlands, pp. 193-207). 

As a conclusion this book is a good introduction for Masters and 

Ph. D students, teachers, researchers and natural-resource managers 

to the theory and practices of ecological restoration. 

Cartografia geobotanica 

F. Pedrotti 

Pitagora editrice, bologna: 236 p. (2004). 

email : pited@pitagoragroup.it 

Ce livre, rédigé par Franco Pedrotti, professeur à l’Université de 

Camerino (Italie) se propose de dresser un panorama des différentes 

méthodes et techniques utilisables pour la cartographie de la végétation. 

Il repose sur le fruit d’une longue expérience de l’auteur, dans le 

cadre de la spatialisation des données de la flore et de la végétation, et 

de très nombreux exemples sont donc issus de ses travaux antérieurs. 

Ce traité présente la cartographie des divers niveaux d’organisation de 

la flore et de la végétation, depuis l’espèce (individus et populations) 

jusqu’aux unités phytogéographiques supérieures. 

L’ouvrage est divisé en 14 chapitres, d’inégale importance 

(le chapitre 3, consacré à la « cartographie synusiale » ne comporte 

qu’une seule page…). L’auteur débute par un chapitre introductif 

présentant la cartographie géobotanique, puis développe les 

aspects cartographiques au niveau des populations (chapitre 2) et 

de l’ensemble de l’aire de distribution d’une espèce (chapitre 4). 

Les quatre chapitres suivants sont consacrés à la cartographie de 

la végétation, et de nombreux exemples de cartes sont fournis. Le 

chapitre 9 est dévolu à la cartographie phytogéographique, avec 

quelques exemples de cartes de subdivision phytogéographique 

(Monde, Espagne, Italie). Les derniers chapitres concernent plus 

particulièrement les apports de la cartographie géobotanique à la 

gestion environnementale et à la conservation des territoires. Une 

bibliographie assez riche termine cet ouvrage. 

Au final, il s’agit d’un travail descriptif et qui aborde les 

thèmes classiques de la géobotanique, souvent sur une base 

phytosociologique, sans réellement proposer de méthodologies 

claires ; sur quelles bases, par exemple, les cartes de subdivisions 

phytogéographiques (chapitre 9) s’opèrent-elles ? Et le traitement 

de la cartographie de la biodiversité végétale en trois pages est 

bien décevant, alors que de nombreux travaux récents existent ! 

Le plus dommageable est l’absence de prise en compte des 

développements récents issus de la modélisation biogéographique 

(analyses DIVA, Parsimony Analysis of Endemicity…) qui ont 

permis des avancées significatives dans ce domaine, sans parler 

de la phylogéographie qui n’est même pas évoquée… Il manque 

en fait clairement tout le volet concernant la simulation de la 

répartition géographique des espèces et de la végétation ; cette 

thématique est pourtant en plein essor (voir A. Guisan, 2003, 

Saussurea 33, pour une synthèse récente en français), notamment 

dans le cadre des études de la réponse biotique face aux 

changements globaux. 

THIERRY DUTOIT 

UMR INRA-UAPV 406 

Écologie des invertébrés 

IUT, Université d’Avignon et des Pays de Vaucluse 

FRÉDÉRIC MÉDAIL 

Institut méditerranéen d’écologie et de paléoécologie 

(IMEP, UMR CNRS 6116), 

Université Paul-Cézanne-Aix-Marseille III. 


Résumés de thèse 

Carey M. Suehs 

Facteurs écologiques et évolutifs influençant 

les processus d’invasion chez Carpobrotus (Aizoaceae) 

en région méditerranéenne 

Thèse de doctorat en sciences soutenue le 8 juillet 2005 

à l’université Paul-Cézanne / Aix-Marseille III, 

Institut méditerranéen d’écologie et de paléoécologie 

(IMEP, UMR-CNRS 6116) 

Jury 

Frédéric Médail (maître de conférences, université Paul-Cézanne, directeur 

de thèse), Laurence Affre (maître de conférences, université Paul-Cézanne, 

co-directrice de thèse), Carla M. d’Antonio (professeur, rapporteur), John 

Thompson (directeur de recherche, rapporteur), Serge Muller (professeur, 

examinateur), Jacques Maillet (professeur, examinateur). 

Malgré la présence de Carpobrotus edulis et C. affine acinaciformis 

sur les côtes méditerranéennes françaises depuis 200 ans, peu 

d’études ont abordé la dynamique d’invasion de ces plantes grasses 

d’origine sud-africaine. La prise en compte à la fois du rôle 

des caractères génétiques et reproducteurs, de la forte fréquence 

d’hybridation, des pollinisateurs indigènes et de la dissémination 

des graines par des mammifères introduits en liaison avec les phénomènes 

liés à l’insularité, suggère que l’invasion des Carpobrotus 

spp. constitue un phénomène complexe et évolutif qui a pour 

conséquence des changements profonds au sein des écosystèmes 

insulaires envahis. Ceci résulte, notamment, en une diminution du 

pourcentage d’abondance de divers groupes d’espèces végétales, 

en une transformation du sol dans le cas de C. affine acinaciformis, 

en une restructuration des réseaux des pollinisateurs et en une 

facilitation des mammifères introduits. La plus forte fréquence 

d’occurrence de C. edulis dans le sud-est de la France par rapport 

à C. affine acinaciformis est potentiellement expliquée par des 

différences génétiques et reproductrices entre ces deux taxons. 

Plusieurs indications d’hybridations fréquentes pourraient expliquer 

une partie de la nature envahissante de ces deux xénophytes, 

leurs changements par rapport à leur aire d’origine, et leur fort 

potentiel évolutif avec possibilité de polyploïdisation. L’ensemble 

de ces résultats suggère que l’étude de la biologie évolutive des 

espèces envahissantes est nécessaire pour mieux comprendre les 

processus d’invasion et l’évolution contemporaine. 

Mots-clés : Carpobrotus, endozoochorie, évolution, invasion biologique, 

hybridation, pollinisateur, structure génétique, système 

de reproduction, xénophyte. 

Buisson Elise 

Ecological restoration of Mediterranean plant 

communities: case studies in Southern France 

and Coastal California (USA) 

Thèse de doctorat en sciences soutenue le 20 septembre 2005 

à l’université Paul-Cézanne / Aix-Marseille III, Institut méditerranéen 

d’écologie et de paléoécologie (IMEP, UMR-CNRS 6116) 

Jury 

Thierry Dutoit (professeur, université d’Avignon), Emmanuel Corcket 

(MCF, université Bordeaux 1), directeur et co-directeur ; Karen Holl 

(professeur, university Santa Cruz, California, USA), Richard Michalet 

(professeur, université Bordeaux 1), rapporteurs ; Thierry Tatoni (professeur, 

université Paul-Cézanne), James Aronson (directeur de recherche, 

CNRS-CEFE Montpellier), examinateurs ; Sean Anderson (docteur, 

Stanford University), Grey Hayes (docteur, Elkhorn Slough National 

Estuarine Research Reserve, USA), membres invités. 

Disturbances, events that change the structure of an entity 

(ecosystem, community, population) by modifying resources, 

substrates or physical environments, are a major factor in entity 

dynamics. Resilience is the ability of the entity to resist or recover 

from an exogenous disturbance (originating from outside forces 

while an endogenous disturbance originates within the entity 

and is thus a part of its “normal” functioning). Disturbances are 

studied because they are at the heart of theories in community 

ecology. They maintain spatial heterogeneity and biodiversity 

dynamics but the mechanisms through which they function 

have yet to be more comprehensively described. The concept of 

resilience is important because it is linked to that of stability and 

to (functional) biodiversity. Its study helps understanding the 

relationships between disturbance, biodiversity and ecosystem 

functioning. 

The absence or low long-term resilience of an ecosystem to an 

exogenous disturbance suggests that irreversibility thresholds have 

been crossed and that the ecosystem is dysfunctional. The second 

objective of this thesis is to use such an opportunity to isolate one 

or several functions and understand better the whole functioning 

of the ecosystem. Grassland species are introduced to degraded 

ecosystems as functions are manipulated (environmental factors 

such as soil fertility and stone cover and biotic factors such as 

grazing, competition/facilitation of neighbor species). A factorial 

experiment allows for the isolation of each of the tested functions 

and for the understanding of their combined effect. Both experiments 

were carried out in herbaceous ecosystems as they are a 

good study model. They are widely distributed and species-rich. 

They have been subjected to endogenous disturbances (e.g. graz- 

111 


◆ RÉSUMÉS DE THÈSES 

112 

ing) under which they were formed and to exogenous disturbances 

under which they have recently been altered (plowing in 

the 19 or 20 th century). Both types of disturbances are relatively 

remnant. Mediterranean ecosystems are particularly interesting 

as they are species-rich (local hot-spot of biodiversity) and have 

a wide variety of endogenous and exogenous disturbances. Our 

study sites thus included several Mediterranean grasslands in 

the plain of La Crau (South-Eastern France) and in California 

Central Coastal. Comparing similar models subjected to different 

conditions and histories allows a better understanding of some 

basic processes. 

In La Crau (South-Eastern France), melon cultivation (1965- 

1985) followed by cereal and alfafa cultivation greatly disturbed the 

steppe vegetation. To assess the resilience of the steppe vegetation, 

we studied the margins of three abandoned melon fields, all adjacent 

to the same remnant of steppe. Data on soil, above-ground vegetation, 

seed bank, seed rain and seed dispersal by ants showed that 

resilience is low. This suggests that the presence of remnant patches 

of steppe do not play a role in propagule dissemination and that the 

ecosystem is dysfunctional. In order to better understand the functioning 

of the steppe, four ecosystem functions were manipulated 

in a factorial experiment: soil fertility, sheep grazing, stone percent 

cover and arable weed interactions. To do so, Thymus vulgaris and 

Brachypodium retusum identified as keystone species (perennial 

and dominant plant of the undisturbed steppe) were sown and 

transplanted on three plots on a fertility gradient: an abandoned 

melon field, an abandoned cereal field and a remnant patch of 

steppe. Results show that nutrient-rich/mycorrhizae-poor soil is 

not suitable for both steppe species (stress tolerant) to germinate. 

However, it is suitable to their establishment and growth if they 

are planted. Grazing strongly reduces seedling growth. Tall neighbors 

compete against both species for light. Shorter neighbors 

(cereal field) do not compete as much against Thymus. Restored 

stone percent cover plays a greater positive role on focal species 

growth where neighbors are short (cereal field). Results will be 

applied to ecological restoration. 

In coastal California, grasslands have been degraded by 

agriculture and over-grazing. There is no autogenic restoration, 

partly, because the reference ecosystem (coastal prairie) and seed 

sources are rare, and because exotic grasses and forbs are particularly 

invasive. Re-introduction of grazing alone does not make the 

ecosystem more resilient (grazing has a positive effect on native 

annual forbs, no effect on native sedges and grasses and a negative 

effect on native perennial forbs). In order to better understand 

the coastal grassland functioning, three ecosystem functions were 

manipulated in a factorial experiment: grazing, topsoil and exotic 

interactions. To do so, two keystone species (native Californian 

perennial species) were sown and transplanted on three sites: two 

sites for Danthonia california within seven km from the coast 

and two site for Nassella pulchra 25 km from the coast. Results 

show that heavy cattle grazing strongly reduces seedling growth, 

while light grazing (rabbits, ground squirrels and deer) may help 

it. However, within grazed plots, cattle decreases the negative 

effect of neighbors. Topsoil removal may not be necessary for 

plant establishment but may be useful in the long run. Results are 

discussed within the framework of ecological restoration. 

For both locations, disturbances and resilience have been identified 

and degraded functions studied. A better understanding of 

ecosystem functioning shows, at an applied level, the functions 

that need to be manipulated to successfully re-introduce dominant 

grassland species to degraded ecosystems. Whether high biodiversity 

leads to better ecosystem functioning is still being discussed. 

However, many ecosystem functions may only depend on a few 

dominant species. One efficient ecological restoration strategy 

is to first improve habitats in order to restore major functions. 

Then, dominant species of the reference ecosystem (nurse plants) 

should be re-introduce in order to restore some other functions 

and facilitate the re-establishment of the whole biodiversity (e.g. 

plant and insects). Such a future study would be an opportunity 

to assess the role of positive biotic interactions (facilitation) in 

improving biodiversity within stressful conditions as well as the 

role of biodiversity on ecosystem functioning. 

Key-words: disturbance, resilience, plant dispersal, competition, 

seed bank, seed rain, myrmecochory, competition, facilitation, 

plant community, edge effect, ecological restoration, sheep grazing, 

dry grasslands, Mediterranean areas, California, Southern 

France. 

Ricouart Francine 

La gestion de l’espace et la prévention des incendies 

de forêt dans les Pyrénées méditerranéennes : paysage, 

diversité des Rhopalocères et réduction de combustible 

Thèse de doctorat en géographie soutenue le 19 septembre 2005 

à l’université de Toulouse-Le Mirail, UMR-CNRS 5602 GEODE 

(géographie de l’environnement), UFR sciences, espaces, sociétés, 

Département de géographie aménagement, Maison de la recherche, 

5 allées Antonio-Machado, 31058 Toulouse cedex 01. 

Jury 

Thierry Dutoit (professeur, université d’Avignon), Jean-Pierre Lumaret 

(professeur, université Montpellier), rapporteurs ; Jean-Charles Filleron 

(professeur, université Toulouse II), Jean-François Galtié (chargé de 

recherche CNRS, Toulouse), examinateurs ; Jean-Paul Métailié (directeur 

de recherche CNRS, Toulouse), président ; Luc Legal (maître de 

conférence, université Paul-Sabatier, Toulouse), directeur. 

Ce travail s’inscrit dans une problématique multiple qui 

traite de la question des incendies, de la gestion de l’espace et de 

l’évaluation des impacts des coupures de combustible. Il vise à 

mieux comprendre les impacts des aménagements réalisés pour 

la gestion des milieux méditerranéens dans un but de DFCI 

(Défense des forêts contre les incendies), tant au niveau de la 

dynamique paysagère qu’au niveau des éventuels marqueurs des 

modifications introduites par ces nouveaux milieux. Il vise aussi 

à proposer des outils aux acteurs locaux de l’aménagement pas- 


RÉSUMÉS DE THÈSES ◆ 

toral et de la DFCI pour l’élaboration des aménagements futurs. 

Cette recherches s’appuie sur une analyse globale des milieux et 

sur une approche fine des formations végétales et de la faune de 

papillons. Les impacts paysagers et écologiques des réductions 

de combustibles sont appréciés à partir d’exemples localisés dans 

les Pyrénées-Orientales, représentatifs des milieux rencontrés en 

région forestière et sub-forestière méditerranéenne. Pour mesurer 

ces impacts, nous avons étudié la diversité biologique et nous 

avons choisi d’utiliser en particulier les Lépidoptères « diurnes » 

(Rhopalocères). Notre propos est ainsi essentiellement basé sur 

la mise en relation de ces populations avec les milieux évoqués. 

Pour réaliser une approche originale de classification hiérarchique 

selon les sites où les espèces sont relevées, ceux-ci ont été classés 

selon deux critères écologiques, niveau d’ouverture du milieu 

et caractéristiques hygrométriques. À partir de la matrice ainsi 

constituée, nous obtenons une répartition théorique des espèces 

en six groupes fonctionnels, marqueurs théoriques de six types 

de milieu. Cette approche originale de classification hiérarchique 

appliquée aux relevés effectués indique des assemblages d’espèces, 

selon les zones, très proches du regroupement théorique. 

Par contre, la réorganisation obtenue dans les arbres par classe 

d’abondance ne nous permet pas d’affirmer que nous obtenons 

une structure stable, car les espèces, pour une classe donnée, sont 

trop peu nombreuses. La cartographie des Rhopalocères permet 

de mettre en relation les lieux, les dates et l’abondance des espèces 

inventoriées. Cela nous permet également de visualiser rapidement 

les différences de faune entre les sites et entre les zones. 

Pour rechercher les facteurs qui établissent cette répartition, 

nous avons réalisé une analyses discriminante par une analyse 

factorielle, multiple puis une classification des facteurs. La correspondance 

des Rhopalocères avec les plantes-hôtes relevées est 

réalisée par une analyse factorielle des correspondances binaires 

sur la matrice mettant en relation Rhopalocères et plantes (198 

espèces végétales et 81 espèces de Rhopalocères en individus) 

suivie d’une classification hiérarchique. 

Nos résultats montrent que les aménagements de prévention 

entraînent des modifications de la faune locale. Mais, ces changements 

sont à mettre en relation avec l’échelle des aménagements. 

La faune totale s’est enrichie d’espèces de milieux ouverts et 

anthropisés. Les acquis de trois années de suivi ont permis de 

valider l’utilisation des Lépidoptères comme marqueurs des 

changements observés dans le cadre de la prévention des grands 

incendies de végétation. Les résultats apportés par ce travail 

sont encourageants quant à l’utilisation des Rhopalocères dans 

les études des impacts des mesures de gestion de l’espace sur la 

diversité biologique. Il serait maintenant important de poursuivre 

cette étude à long terme, incluant plusieurs cycles d’entretien 

(ouverture/fermeture) dont le brûlage dirigé, et en diversifiant les 

milieux suivis. Il faut que la base de données établie soit complétée 

par d’autres relevés suivant le même protocole. Une comparaison 

pourrait être faite entre les impacts des débroussaillements et ceux 

des incendies sur les populations de Rhopalocères. 

Mots-clés : déprise agro-pastorale, analyse structurale, bio-indicateurs, 

biodiversité, enfrichement, reboisement, politiques de gestion, 

analyses multi-variées, Pyrénées-Orientales, Albères, Aspres, 

Fenouillèdes. 

113 


ÉDISUD 

ÉCOLOGIE ❈ NATURE 

REVUE ECOLOGIA MEDITERRANEA 

Numéros thématiques 

Biologie de la conservation 

et des espaces naturels en Crau 

vol. 30 (1/2), 132 p., 2004, 31 € 

The sustainability of chesnut forest 

in the mediterranean region 

vol. 26 (1/2), 179 p., 2000, 25 € 

Écologie et conservation des mares 

temporaires méditerranéennes 

vol. 21 (2), 135 p., 1998, 25 € 

Connaissance et conservation 

de la flore des îles de la Méditerranée 

vol. 21 (1/2), 379 p., 1995, 25 € 

Volume jubilaire dédié 

au professeur Pierre Quézel 

vol. 16, 459 p., 1990, 20 € 

Offre spéciale sur les 4 volumes suivants : 

30,49 € ‡ 15 € 

Influence of fire on the stability 

of Mediterranean forest ecosystems 

vol. 13 (4), 196 p., 1987, 15 € 

Changements hydrologiques 

de la zone tempérée au cours 

des quinze derniers millénaires 

vol. 11 (1), 236 p., 1985, 15 € 

Définition et localisation 

des écosystèmes méditerranéens 

terrestres 

vol. 8 (1/2), 495 p., 1982, 15 € 

Études écologiques 

et ethnologiques dans le Niolu 

vol. 6, 232 p., 1980, 15 € 

PAYSAGE 

Sainte-Victoire 

À la découverte d’un sentier 

Association pour le reboisement 

et la protection du Cengle - Sainte-Victoire 

Découvrir le massif de Sainte-Victoire 

par son versant sud. 

22,5 x 20 cm • 144 p.• © 1998 • broché 

• tout en couleur • 20,58 € • ISBN 2-7449-0012-5 

Paysages de terrasses 

R. Ambroise, P. Frapa, S. Giorgis 

Photographies de V. Motte et R. Sauvaire 

Très beau livre illustré, hommage au travail des paysans 

qui ont transformé une nature parfois hostile 

en d’immenses jardins cultivés. 

24 x 30 cm • 192 p. • © 1989 • relié • nbr. photos N&B 

et couleur • 28,97 € • ISBN 2-85744-449-8 

FLORE 

La Connaissance des Succulentes 

et Xérophytes du Monde 

Francis et Martine Bugaret 

Un ouvrage de référence fondamental 

par d’éminents spécialistes. 

Coll. “La Connaissance de…” • 17 x 24 cm • 408 p. 

• © 2006 • cartonné • très nombr. ill. couleur • 33 € 

• ISBN 2-7449-0083-4 

La Végétation de la Corse 

Jacques Gamisans 

Un ouvrage de référence fondamental par le plus 

éminent spécialiste actuel sur la question. 

16 x 24 cm • 408 p. • © 1999 - 2003 • broché • ill. N & B 

et 16 p. couleur • 24,39 € • ISBN 2-7449-0083-4 

La Flore endémique de la Corse 

Jacques Gamisans – Jean-François Marzocchi 

Chaque espèce, identifiée par une photographie, 

est décrite avec ses noms latin, corse et français, 

l’indication de la famille, la période de floraison, 

son aire géographique, son indice de rareté, 

son écologie et sa répartition en Corse. 

Une liste exhaustive figure en fin d’ouvrage. 

16 x 22,5 cm • 192 p. • © 1996-2003 • broché avec rabats 

• 14,94 € • ISBN 2-85744-777-9 

La Flore des montagnes méditerranéennes 

Christian Boucher 

De l’Andalousie à l’Anatolie et au Proche-Orient, 

250 plantes décrites et illustrées. 

16 x 22,5 cm • 208 p. • © 2000 • 256 photos couleur 

• 3 cartes • broché avec rabats • 24,39 € 

• ISBN 2-7449-0146-6 

Flore du Verdon 

Un parc en fleurs 

Parc naturel régional du Verdon – Georges 

Guende 

Après présentation du parc, l’auteur décline, par secteur 

géographique, les espèces qui peuplent cette région. 

16 x 22,5 cm • 192 p. • © 2004 • broché • ill. couleur 

• 20 € • ISBN 2-7449-0467-8 

Flore du Luberon 

Georges Guende 

Nouvelle édition. Encore relativement bien préservé, 

le Luberon offre à sa flore une grande diversité de climat et 

de types d’habitat. 

Coll. “Luberon Images et Signes” • 16,5 x 22,5 cm 

• 144 p. • © 1993-2004 •broché • nombr. ill. couleur 

• 15 € • ISBN 2-7449-0476-7 

Orchidées sauvages du Luberon 

Roland Martin 

Le Luberon, en Provence, offre aux orchidées un habitat 

et un climat privilégiés. Découvrir ces trésors 

de la nature est un enchantement. 


• 224 p. • © 2005 •broché • env. 200 photographies couleur 

• 15 € • ISBN 2-7449-0468-6 

Fleurs de Vanoise 

Parc national de la Vanoise 

Un guide qui recense 615 espèces florales du massif 

de la Vanoise, dont 500 illustrées par des photographies 

couleur. 

12 x 18 cm • 320 p. • © 1993 • cartonné 

• 19,81 € • ISBN 2-85744-653-5 

Splendeur et harmonie des plantes libres 

Pierre Vignes 

Grâce à de magnifiques planches photographiques, 

inhabituelles et originales, démarche scientifique 

et sens artistique s’allient à la découverte esthétique. 

22 x 30 cm • 160 p. • © 2004 

• 28,50 € • ISBN 2-7449-0450-3

Herbier vagabond 

Claude Meslay 

Aquarelles de M.-F. Delarozière 

Une initiation variée et passionnante 

à la promenade botanique. 

22,5 x 26,5 cm • 160 p. • © 2002 • cartonné 

• aquarelles • 29 € • ISBN 2-7449-0331-0 

Herbier des bords de l’eau 

Claude Meslay 

Aquarelles de M.-F. Delarozière 

150 plantes des lacs, étangs, mares, ruisseaux, canaux 

d’irrigation et leur voisinage immédiat. Un regard sur les 

zones aquatiques à la flore si riche. 

22,5 x 26,5 cm • 176 p. • © 2004 • cartonné 

• aquarelles • 29 € • ISBN 2-7449-0387-6 

FAUNE 

Le Thon de Méditerranée 

La saga d’un poisson mythique 

Patrick Mouton 

Le thon de Méditerranée effectue chaque année de 

grandes migrations de plusieurs milliers d’individus. 

Une saga passionnante illustrée d’une iconographie 

abondante et variée. 

Coll. “Mémoire de mer” • 27 x 21 cm • 144 p. © 2004 

• cartonné sous jaquette • photos couleur et N&B 

• 26 € • ISBN 2-7449-0408-2 

Vie secrète au royaume montagne 

Les quatre saisons de la faune alpine 

Brys Bonnal – Préface Allain Bougrain Dubourg 

Saison après saison, des images sans équivalent, 

mais aussi un texte alliant avec beaucoup de finesse 

les atmosphères poétiques et les faits scientifiques. 

24 x 23 cm • 144 p. • © 1997 • cartonné • photos couleur 

• 22,10 € • ISBN 2-85744-917-8 

Faune du Luberon 

Max Gallardo 

Le Luberon offre à sa faune une grande diversité de 

climat et de types d’habitat. Cet ouvrage permet d’en 

aborder la connaissance selon ses milieux les plus typés. 

Coll. “Luberon Images et signes” • 16,5 x 22,5 cm • 144 p. 

• © 1993 • broché • 13,73 € • ISBN 2-85744-663-2 

Le Luberon des insectes 

Parc naturel régional du Luberon 

Un aperçu des 17 000 espèces d’insectes répertoriées dans 

le Luberon. Environ 150 d’entre elles, parmi les plus 

significatives, sont citées, dont une centaine illustrées de 

dessins, d’aquarelles et de gouaches. 


• 120 p. • © 1998 • broché • nbr. ill. couleur 

• 12,96 € • ISBN 2-7449-0058-3 

Garrigue secrète 

Patrick Lorne 

D’innombrables instants de vie des insectes de la 

garrigue, rythmés selon les quatre saisons. Chaque 

partie est introduite par un texte dans lequel l’auteur 

dévoile les mœurs de cette micro-faune et ses propres 

observations. 

26 x 23 cm • 144 p. • © 1998 • relié sous jaquette 

• tout en couleur • 22 € • ISBN 2-7449-0035-4 

Les Insectes, amis de nos jardins 

Pourquoi et comment les protéger 

Vincent Albouy, Jean-Paul Delfino 

Coll. “Que la nature est belle !” • 17 x 23 cm • 136 p. 

• © 1995 • broché • 13,73 € • ISBN 2-85744-776-0 

L’ARBRE 

L’Arbre familier en Provence 

De la vocation du platane et de quelques autres arbres 

Annie-Hélène Dufour 

Photos Martin Johansen et Annie-Hélène Dufour 

Au travers de cet aperçu historique, le lecteur découvrira 

si le platane est réellement devenu l’élément végétal 

dominant, et quelle place il occupe dans les pratiques et 

les représentations des Provençaux. 

16 x 24 cm • 112 p. • © 2001 • broché • photos couleur 

et N&B • 15,50 € • ISBN 2-7449-0296-9 

Le Livre de l’Olivier 

Marie-Claire Amouretti, Georges Comet 

Les aspects culturels, économiques et culturaux de 

l’olivier en Méditerranée. 

224 p. • © 1999 • 15 € • ISBN 2-7449-0198-9-7 

Liège de Méditerranée 

Gérard Dessain, Margaret Tondelier 

Le liège, ses mythes, ses utilisations. Un livre où 

découvrir origines, transformations, utilisations passées 

ou présentes du liège. 

21,5 x 29 cm • 80 p. • © 1991 • cartonné 

• 14,94 € • ISBN 2-85744-548-2 

Le Tilleul 

Arbre d’amour, arbre de liberté 

Guy Bontempelli 

Le point sur les croyances, superstitions, culture, 

utilisation du bois, vertus thérapeutiques de cet arbre 

symbole. 

24 x 27,5 cm • 168 p. • © 1989 • relié • 22,87 € 

• ISBN 2-907836-00-5 

Le Noyer et la noix 

J.-J. de Corcelles, J.-Y. Catherin et Robert Mazin 

Ce livre répond à toutes les questions que l’on peut se 

poser sur cet arbre et sur ce fruit secret et méconnu. 

160 p. • © 1995 • 14,94 € • ISBN 2-85744-791-4 

Châtaignes et châtaigniers 

en régions méditerranéennes françaises 

G. Briane, C. Catoire, P. Challaye, J.-J. Gianni, 

R. et A. Sauvezon 

Tous les usages propres au châtaignier et à ses fruits, 

comment on peut sauver, aujourd’hui, le patrimoine 

qu’il constitue, en vivre, l’aimer. 

168 p.• © 2001 • 15,10 € • ISBN 2-7449-0186-5 

LA FORÊT 

La Futaie irrégulière 

Théorie et pratique de la sylviculture irrégulière, 

continue et proche de la nature 

Max Bruciamacchie, Brice de Turckheim 

Max Bruciamacchie, enseignant-chercheur à l’Engref, 

Nancy, et Brice de Turckheim, praticien forestier, 

associent ici leurs compétences pour présenter aux 

praticiens sylvicoles et à tous ceux qui s’intéressent 

au devenir de nos forêts, ce livre destiné à guider leur 

réflexion. 

17 x 24 cm • broché • N&B et couleur • 288 p. 

• 29,50 € • ISBN 2-85744-304-8 

Incendies de forêt et argent public 

Paul-Henry Fleur 

Quelles interventions ? Quelle prévention ? 

Quels résultats ? Une enquête objective et sérieuse. 

192 p. • © 2004 • 18 € • ISBN 2-7449-0484-8 

L’Économie de la forêt 

Mieux exploiter un patrimoine 

Henri Prévôt 

Une nouvelle approche qui embrasse tous les aspects 

de la forêt. L’auteur explique le blocage dont souffre ce 

secteur économique. Puis, il propose de nouveaux outils 

financiers et de nouvelles relations commerciales. 

16 x 24 cm • 176 p. • © 1993 • broché 

• 19,52 € • ISBN 2-85744-651-9 

ÉDITIONS ÉDISUD – 3120, route d’Avignon – FR 13090 Aix-en-Provence 

Tél. [0033](0)4 42 21 61 44 – Fax [0033](0)4 42 21 56 20 – info@edisud.com – www.edisud.com

ecologia mediterranea - UniversitÃ© d'Avignon et des Pays de Vaucluse

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?