MEDITERRANEAN TEMPORARY PONDS

The role of spatial environmental factors as determinantsof large branchiopod distribution in Tunisian temporaryponds

Fabio Stoch . Michael Korn . Souad Turki .

Luigi Naselli-Flores . Federico Marrone

Received: 17 October 2015 / Revised: 5 December 2015 / Accepted: 26 December 2015 / Published online: 24 February 2016

� Springer International Publishing Switzerland 2016

Abstract The influence of spatial and environmental

factors in explaining the structure of large bran-

chiopod assemblages at different spatial scales is still

poorly explored. We hypothesized that the extent of

actual spatial connectivity, and thus the spatial distri-

bution of a metacommunity, may depend on the

environmental conditions as represented by climatic

gradients and the structural characteristics of the

landscape. To test this hypothesis, the distributional

patterns of 14 large branchiopod species in a set of 177

temporary water bodies repeatedly sampled across

Tunisia and on its main islands were analysed.

Physical, chemical, morphological and climatic char-

acteristics of the studied water bodies were collected

as well, and spatial structures were described using

distance-based Moran’s Eigenvector Maps. Distance-

based Redundancy Analysis and variance partitioning

explained more than one-half of total variation.

Mantel’s autocorrelograms demonstrated that species

composition was spatially autocorrelated at shorter

distances in the Mediterranean part of the country than

in the southern, more arid part. These results suggest

that both dispersal limitation and species response to

spatially structured environmental gradients might be

involved in determining large branchiopod distribu-

tion in Tunisia and that these patterns may greatly vary

on a regional basis.

Keywords Species distribution � Climatic

gradients � Environmental filters � Geographicaldistances � Spatial scale

Guest editors: Simonetta Bagella, Dani Boix, Rossella

Filigheddu, Stephanie Gascon, Annalena Cogoni /

Mediterranean Temporary Ponds

Electronic supplementary material The online version ofthis article (doi:10.1007/s10750-015-2637-y) contains supple-mentary material, which is available to authorized users.

F. Stoch

Department of Life, Health & Environmental Sciences,

University of L’Aquila, Via Vetoio, Coppito,

67100 L’Aquila, Italy

M. Korn

DNA-Laboratory, Museum of Zoology, Senckenberg

Natural History Collections Dresden, Konigsbrucker

Landstrasse 159, 01109 Dresden, Germany

S. Turki

National Institute of Marine Sciences and Technologies,

Rue du 02 mars 1934 28, 2025 Salammbo, Tunisia

L. Naselli-Flores

Department of Biological, Chemical and Pharmaceutical

Sciences and Technologies, University of Palermo, via

Archirafi 28, 90123 Palermo, Italy

F. Marrone (&)

Department of Biological, Chemical and Pharmaceutical

Sciences and Technologies, University of Palermo, via

Archirafi 18, 90123 Palermo, Italy

e-mail: federico.marrone@unipa.it

123

Hydrobiologia (2016) 782:37–51

DOI 10.1007/s10750-015-2637-y

Introduction

Temporary ponds and pools are the most representa-

tive natural water bodies in arid and semi-arid

countries, where hot and dry conditions restrict most

aquatic systems to seasonality (Nhiwatiwa et al.,

2011). In the Mediterranean region, they are probably

the aquatic ecosystems that achieve the most remark-

able biodiversity due to their very peculiar aquatic

biota which is able to bridge the dry periods and to deal

with their extreme variability (Williams, 2006;

Naselli-Flores & Barone, 2012). At the same time,

temporary ponds are also the most threatened aquatic

ecosystems in the Mediterranean area and they are

presently disappearing at a very fast rate (Zacharias &

Zamparas, 2010).

Large branchiopod orders Anostraca, Notostraca,

Laevicaudata and Spinicaudata include important

flagship species of temporary waters (King, 1998;

Colburn, 2004), possessing characteristics, such as

rapid life cycle and diapause, which enable them to

persist as permanent residents of these ecosystems.

The presence of resting stages suggests an important

role of passive dispersal in explaining large bran-

chiopod distribution in temporary ponds (Champeau

& Thiery, 1990). Moreover, it was recently suggested

that large branchiopods may play a keystone role in

temporary aquatic ecosystems, and shape microcrus-

tacean communities under a wide range of environ-

mental conditions (Waterkeyn et al., 2011).

Unfortunately, human disturbance and destruction of

their often neglected natural habitats caused them to

be globally vulnerable and to undergo a serious

decline of their distribution (Samraoui et al., 2006;

Brendonck et al., 2008; Demeter & Stoicescu, 2008).

For these reasons, the use of large branchiopods in

monitoring temporary ponds is highly recommended

(Waterkeyn et al., 2011).

Although large branchiopods have a worldwide

distribution, their diversity patterns, distribution and

conservation status are poorly known, especially for

large parts of Africa (Brendonck et al., 2008).

Distribution and ecology of North African large

branchiopods are currently best known from Morocco

(Thiery, 1987; Roux & Thiery, 1988, van den Broeck

et al., 2015, and references therein), and Algeria

(Gauthier, 1928a, b; Samraoui & Dumont, 2002;

Samraoui et al., 2006). Conversely, after the seminal

contributes of Gurney (1909) and Gauthier (1928a, b),

and some further data recently published (e.g. Korn

et al., 2006; Turki & Turki, 2010, and references

therein), a comprehensive checklist for Tunisian large

branchiopods is still missing.

Several studies have investigated the importance of

environmental factors and landscape structure in

explaining occurrence patterns of large branchiopods

in an attempt to design more appropriate conservation

strategies (Marrone et al., 2006; Angeler et al., 2008;

Waterkeyn et al., 2008). In the light of the high

potential for passive dispersal of large branchiopods,

the spatial configuration of habitat patches was

recently studied in a system of temporary pans in

Zimbabwe (Nhiwatiwa et al., 2011). However, a

considerable amount (75–81%) of the variation in

species composition among these pans remained

unexplained, suggesting that these investigations are

complicated by (i) the erratic presence of the species in

the study area due to their high dispersal ability and (ii)

the sharp decoupling existing between passive disper-

sal potential and actual establishment rates. The latter

largely depend on the chances that newcomer organ-

isms have to pass through a series of chemical,

physical and biological filters that may prevent their

successful colonization of a new environment (see

Incagnone et al., 2015 for a review). In addition,

structural connectivity among ponds (i.e. the number

and relative distance of ponds in a given area) may

also play a role especially in arid and semi-arid areas,

since it may constitute a spatial limit to the dispersal of

propagules (Naselli-Flores et al., 2016).

Unravelling the relative contribution of local and

regional processes in shaping the diversity and species

distribution patterns is necessary for monitoring

changes in global biodiversity (Gaston & Spicer,

2004) and essential for the development of conserva-

tion strategies of temporary ponds (Waterkeyn et al.,

2011). In this study, we used a dataset of large

branchiopod occurrences in Tunisian temporary water

bodies with the explicit aim to investigate the relative

role of some selected spatial and environmental

variables at different spatial scales in explaining the

composition of species assemblages and their distri-

bution. Actually, spatial scales and the relative

distances among aquatic ecosystems may also con-

found the balance between deterministic (mediated by

environmental factors) and stochastic (mediated by

dispersal) factors. We hypothesized that the ‘‘extent’’

of actual spatial connectivity, and thus the spatial

38 Hydrobiologia (2016) 782:37–51

123

distribution of a metacommunity, may depend on the

environmental conditions as represented by climatic

gradients and the structural characteristics of the

landscape (e.g. vegetated vs open areas) surrounding

the studied temporary ponds.

Methods

Study area

Tunisia, located in eastern Maghreb, encompasses an

area of about 164,000 km2 with about 1,300 km of

coastline. Despite its relatively small area, the country

has a noteworthy environmental diversity due to its

north–south extent and sharply decreasing rainfall

southward. The Dorsal, which is the eastern extension

of the Atlas Mountains, runs across Tunisia in a north-

easterly direction. North of the Dorsal lies the Tell

Plateau, characterized by hills and plains. In the north-

western corner of Tunisia, the Medjerda mountains

delimit a humid coastal area, which is the wettest part

of the country. This area, which borders the northern

Mediterranean coast, has average precipitation values

higher than 500 mm year-1. Along Tunisia’s eastern

Mediterranean coast lies the Sahel, a broadening

coastal plain; inland from the Sahel, between the

Dorsal and a range of hills south of Gafsa, is located

the steppe, while the southern part of the country is

desert. The southern areas are thus represented by the

desert steppe, the Great Eastern Erg (sand-dune

desert), the hamada (rocky desert), and the chotts

(salt-lakes depressions, the lowest one being Chott el

Djerid, -17 m a.s.l.).

The Tunisian climate (Fig. 1) is a hot-summer

Mediterranean climate in the north (following Kop-

pen-Geiger terminology: see Peel et al., 2007), where

winters are generally mild with moderate rainfall and

summers are hot (temperature can exceed 40�C) anddry. The southern part has a desert climate with hot

summers, warm winters, and low annual rainfall.

Average annual rainfall is lower than 500 mm nearly

everywhere in Tunisia, with the significant exception

of the humid north-western area, where average

annual precipitation can reach 1200 mm.

The study area, based on the climatic and physio-

graphic features mentioned above, is here divided into

three ecological zones following Gauthier (1928a),

Dumont et al. (1979), and Turki & El Abed (1999)

(Fig. 2a): (1) north-western humid Mediterranean

area, with mean annual rainfall[500 mm; (2) eastern

dry Mediterranean area and Tell plateau, with mean

annual rainfall between 300 and 500 mm; these areas

are characterized by natural forest, maquis and

rangelands, which become fragile in the southern

zone; (3) arid, steppic and semi-desert area, with mean

annual rainfall between 100 and 300 mm. No sites

were sampled in the Saharan desert area (mean annual

rainfall\100 mm).

Sampling methods and species identification

On the whole, 177 temporary water bodies located

from 0 up to 870 m a.s.l. were visited and sampled in

the Tunisianmainland and on the islands of Djerba and

Kerkennah (Fig. 2a); the majority of the sampled sites

were visited in different seasons and years (from 2004

to 2012) in order to account also for the possible

presence of species with short life-cycles, which might

be easily overlooked with snapshot sampling.

Crustacean sampleswere collected bymeans of three

different nets: a 125-lm mesh-sized towing net was

used in the openwaters; a 200-lmhand net was used for

the benthic and littoral zones, and a further 600-lmhand

net specifically for the adults, which are able to actively

avoid the thicker nets and, especially if present in low

densities, may be easily overlooked using the conven-

tional zooplankton sampling techniques.

Collected samples were fixed in situ in 90%

ethanol, in order to make them available for both

morphological study and molecular analyses. When

large branchiopod nauplii, juveniles and\or immature

specimens were spotted, the collected samples were

duplicated and in part kept alive in the laboratory in

order to let them reach sexual maturity.

Branchiopods were sorted under the stereomicro-

scope, prepared according to Alonso (1996) and

Dumont & Negrea (2002), and then identified accord-

ing to Thiery (1987, 1996), Alonso (1996) and Korn

et al. (2006, 2010, 2013).

Sample-based rarefaction curves (Gotelli & Col-

well, 2001) were calculated to investigate the exhaus-

tiveness of sampling effort. The software EstimateS

version 9.1 (Colwell, 2005) was used; ICE (Incidence-

based estimator), Chao2, and Jackknife 1 were

selected to estimate species richness (Colwell, 2005).

The rarefaction curves of the mean values of unique

occurrences (i.e. species present in a single site) and

Hydrobiologia (2016) 782:37–51 39

123

Fig. 1 Spatial distribution of (a) average monthly maximum temperature, and (b) mean annual precipitation in Tunisia (1975–2000)

Fig. 2 (a) Spatial distribution of sites in the three ecological zones of Tunisia (see text). (b) Minimum spanning tree connecting the

sampling sites where large branchiopods were collected

40 Hydrobiologia (2016) 782:37–51

123

duplicate occurrences (species present in two sites)

were calculated as well. The analysis was performed

using 9999 Monte Carlo permutations without

resampling.

A binary (presence/absence) species composition

matrix was assembled retaining the 108 sites in which

large branchiopods were found and where all the

necessary environmental parameters and variables

were collected.

Environmental variables

Together with large branchiopod sampling, the fol-

lowing local environmental parameters and variables

were collected for each pond: altitude (m a.s.l.), size

(classes), water temperature (�C), electrical conduc-tivity (lS/cm), turbidity (classes), and macrophyte

cover (classes). Water temperature and conductivity

were measured using a hand-held multiprobe. Water

bodies were classified into three size classes referring

to the surface area covered at maximum water storage

(1:\ 100 m2; 2:[ 100,\ 2500 m2; 3:[ 2500 m2).

Three arbitrary qualitative classes were used to

estimate inorganic water turbidity (from 1: crystal-

clear water, to 3: extremely turbid water). Macrophyte

cover was estimated visually observing the pond

surface; the percentage of macrophyte cover was used

as a proxy for the structural complexity of aquatic

vegetation and hence habitat complexity, varying from

0 (low complexity, macrophytes absent) to 3 (high

complexity, &100% macrophyte cover).

Three climatic variables were used as predictors of

species distribution: actual evapotranspiration, air

temperature, and precipitation; respective data were

obtained from the following on-line databases. Aver-

age monthly maximum temperature (�C) and average

monthly precipitation (mm) data were downloaded

from the WorldClim website (www.worldclim.org:

Hijmanset al., 2005), provided on a 30 arc-seconds

(*1 km) resolution grid; the other climatic variables

available from WorldClim website (average monthly

minimum and mean temperature) were collinear with

average monthly maximum temperature and were not

considered for further analyses. Annual actual evap-

otranspiration (AET) data were extracted from the 30

arc-second resolution worldmap released by Trabucco

& Zomer (2010). AET provides a synthetic index of

water–energy dynamics (O’Brien, 2006), a crucial

factor determining the amount of macrophytes

productivity and the hydroperiod length of the ponds.

Climatic values within a grid cell were attributed to

each pond whose barycentre fell within a given grid

cell using Qgis freeware software version 2.8.1 (QGIS

Development Team, 2015).

Variables were standardized to zero mean and unit

variance to account for their different scales of

measurement. Standardizations were implemented in

R package vegan (Oksanen et al., 2011), implemented

in the open source R software version 3.2.0 (R Core

Team, 2015). A forward selection procedure was used

to select the environmental variables which signifi-

cantly explained the variation in the species compo-

sition matrix; water temperature and elevation where

thus discarded from further analyses.

Spatial variables

The location of each pond was determined with a

precision of 10 meters using a GPS device. Coordi-

nates were expressed as decimal degrees, map datum

WGS84.

Distance-based Moran’s Eigenvector Maps

(dbMEMs: Dray et al., 2012; Legendre & Legendre,

2012) were used to describe the spatial structure of the

dataset. dbMEMs represent a spectral decomposition

of the spatial relationships among the study sites,

leading to multiscale analysis in variation partitioning.

They are a class of multiscale ordination methods that

produce orthogonal eigenvectors used to represent

spatial relationships among samplings (Dray et al.,

2012). The first eigenvectors usually describe broad

spatial structures, encompassing the spatial variation

in the whole sampled area, while the last eigenvectors

(with lower eigenvalues) describe fine spatial struc-

tures, which may capture variation at the scale of

sampling sites (Dray et al., 2012; Legendre &

Legendre, 2012). dbMEM eigenvectors were com-

puted using R package PCNM (Legendre et al., 2010)

from a truncated geodesic distance matrix obtained

with R package fields (Furrer et al., 2011). The longest

distance connecting two ponds in the minimum

spanning tree (Fig. 2b) was used as a threshold to

truncate the distance matrix. A forward selection

procedure with double-stopping criteria (Legendre &

Legendre, 2012), performed in R package packfor

(Dray, 2009), was used to only select dbMEMs that

significantly explained the variation in the species

composition matrix. This procedure recovered 6

Hydrobiologia (2016) 782:37–51 41

123

dbMEMs (cumulative R2 = 0.33, probability level

\0.05). All the selected dbMEMs have positive

autocorrelation up to a well-defined spatial scale.

After examining their semi-variograms (Fig. S2) and

spatial distribution (Fig. 3), dbMEM 1 was arbitrarily

classified as a broad-scale ([100 km) variable,

dbMEMs 2, 3, 4 and 5 were considered as medium

scale (between 50 and 100 km) variables, and

dbMEM11 as a fine-scale (\25 km) variable.

Data analysis

The subsequent multivariate analyses used the binary

species composition matrix, the environmental matrix,

and the spatial matrix including the six dbMEMs.

Relationships between large branchiopod assem-

blages and environmental and spatial variables data

were assessed by a distance-based redundancy anal-

ysis (dbRDA: Legendre & Anderson, 1999) applied to

the similarity species matrix (using Jaccard’s index);

partial dbRDAs were used to assess the relationship

between the species matrix and environmental and

spatial variables separately. Analyses were imple-

mented in the R package vegan (Oksanen et al., 2011).

The spatial and environmental variables were then

used in a variation partitioning analysis (Borcard et al.,

1992; Peres Neto et al., 2006) to determine the among-

site variation for the species dissimilarity matrix that

was explained purely by environmental variation

(E | S), spatially structured environmental variation

(E \ S), or purely spatial variation (S | E). Variation

partitioning was applied to dbRDA using the algo-

rithms of Sokol et al. (2014) implemented in R.

AMantel correlogram (Borcard & Legendre, 2012)

was implemented to assess the pattern of spatial

autocorrelation of the species similarity matrix (cal-

culated using Jaccard’s index) to obtain the geograph-

ical distance threshold above which sites are to be

considered no more autocorrelated. The geographical

distance matrix was divided into ten distance classes

d = 1

dbMEM 1 d = 1

dbMEM 2 d = 1

dbMEM 5

d = 1

dbMEM 4 d = 1

dbMEM 3 d = 1

dbMEM 11

Fig. 3 Distance-based

Moran’s eigenvectors maps

(dbMEMs) significantly

correlated with species

distribution matrix (dbMEM

1: broad-scale variable; 2–5

medium scale; 11 fine-scale)

42 Hydrobiologia (2016) 782:37–51

123

using Sturge’s rule (Scott, 2009) to set the range of

pairwise distances in each class (Legendre & Legen-

dre, 2012). Mantel correlation coefficients were cal-

culated at each distance class and tested for

significance with a permutation test (using 9999

permutations) based on a sequential Bonferroni cor-

rection (a = 0.05; Legendre & Legendre, 2012). All

Mantel analyses were performed using the R package

ecodist (Goslee & Urban, 2013).

Results

Species diversity

The occurrence of large branchiopods was registered

in 108 out of the 177 sites studied to date (i.e. 61% of

the sampled sites; Fig. 2a). In the frame of this survey

14 species were collected (Tab. 1).

Species richness estimates using rarefaction curves

showed that all the estimators did not stabilize with

increasing sample size (Fig. S1). Conversely, the

uniques (i.e. the number of species found in a single

site) and duplicates (i.e. the number of species found in

two sites) decreased, suggesting an efficient sampling

effort. Comparing the number of species observed and

estimated, it can be stated that about 88–93% of the

total species richness was collected so far.

Multi-species co-occurrence was often observed;

the species richness per site (average ± standard

deviation and range) was 2.0 ± 1.3 (1–7); usually,

assemblages were composed of one representative of

each of the three orders of large branchiopods that

occur in the study region. However, up to 7 species

(i.e. 1 Notostraca, 2 Spinicaudata, and 4 Anostraca)

were observed syntopically in some sites occurring in

the Medjerda flood plain.

Detailed information on the sampled sites is

provided by Marrone et al. (2016).

Spatial and environmental drivers of diversity

The explained constrained variation of the first two

axes of the distance-based redundancy analysis

(dbRDA: Fig. 4) was 80%, indicating a high, signif-

icant correlation between predictor variables and the

species distribution similarity matrix; unconstrained

explained variation accounted for 43.4% of total

variation.

The plot of sampling sites on the plane defined by

the first two axes of dbRDA, with environmental and

spatial vectors superimposed on the plot, clearly

showed that, at a broad spatial scale, ponds were

grouped into the three ecological zones defined for the

Tunisian territory (Fig. 4a). Ponds located in the

steppic and semi-desert areas (left side of the graph)

are separated from the ones located in the northern

areas, and there was a transition towards the north-

western wet coastal area ponds which were placed on

the right side of the graph. The first axis represented a

large-scale (high correlation with dbMEM1) climatic

gradient characterized by precipitation and maximum

temperature. The second axis represented a medium-

scale (high correlation with dbMEM 5, and a lower

one with dbMEM4) gradient of conductivity, which

separated within each ecological zone large-sized

ponds characterized by high conductivity values (in

the higher part of the graph) from the smaller ones with

low conductivity but high turbidity or high macro-

phyte cover. Some medium-scale dbMEMs (2 and 3)

and the smaller scale dbMEM11, although showing a

significant positive correlation with the first axis, are

decoupled from environmental variables, while char-

acterizing the non-saline steppic ponds.

Large branchiopod species projected on the plane

defined by the first two dbRDA axes (Fig. 4b) using

their multiple correlation coefficients with the axes

clearly showed that (i) Branchipus schaefferi Fischer,

1834, Streptocephalus torvicornis (Waga, 1842), and

Triops granarius (Lucas, 1864) were significantly

linked to the steppic area, and arranged on the left side

of the first axis; (ii) Chirocephalus diaphanus Prevost,

1803 and Lepidurus apus lubbocki Brauer, 1873

characterized the north-western humid Mediterranean

temporary ponds, and were arranged along the posi-

tive, right side of first axis; finally, Phallocryptus

spinosus (Milne-Edwards, 1840), Artemia salina

(Linnaeus, 1758) (together with the rare Branchinec-

tella media Schmankewitsch, 1873) were exclusive of

saline sites, and located in the upper part of the graph.

Variation partitioning using partial-dbRDAs (ex-

plained variance 56%, Fig. 5) showed the pure and

shared influence of ecological and spatial variables on

large branchiopod species composition. Both environ-

mental variables and the spatial distribution of ponds

affected large branchiopod distribution in Tunisia.

Pure environmental variation (E | S) accounted for 17%

and pure spatial variation (S | E) for 15% of total

Hydrobiologia (2016) 782:37–51 43

123

variation. A large part of environmental variables was

spatially structured, joint variation (E \ S) accounting

for 24% of total variation. Approximately 44% of total

variation remained unexplained.

Mantel correlogram of the species distribution

similarity matrix (Fig. 6) showed a roughly linear

relationship at closer distances. Species composition

showed significant, positive spatial autocorrelation

in the first two distance classes (up to about 100 km:

P\ 0.001 and P\ 0.005, respectively) and nega-

tive autocorrelation (P\ 0.005) in class 4. Space

was unimportant at farther distances (class 5). This

indicated that ponds separated by 100 km or more

significantly differ in species composition. This

large-scale analysis confirmed the results reported

by the first axis of dbRDA analysis, suggesting that

this broad-scale gradient was linked with the

spatially structured climatic variables. The latitudi-

nal climatic gradient was sharp, clearly separating

the fauna of northern (ecological zones 1 and 2:

Mediterranean area and Tell plateau) and southern

(ecological zone 3: steppic and semi-desert areas)

areas. For this reason, Mantel correlograms were

inspected separately for species similarity matrices

of northern and southern sites (Fig. S3). In the

northern area, the correlogram showed a sudden

decrease in autocorrelation which was significant

(P\ 0.05) only at the smallest distance class (i.e.

Table 1 List of collected

large branchiopod species.

For each species, the

number of collecting sites

and the acronyms used in

figures are reported

Species and taxonomical hierarchy Number of sites Acronyms

Anostraca

Family Artemiidae

Artemia salina (Linnaeus, 1758) 3 Asal

Family Branchinectidae

Branchinecta ferox (Milne-Edwards, 1840) 1 Bfer

Family Branchipodidae

Branchipus schaefferi Fischer, 1834 31 Bsch

Family Chirocephalidae

Subfamily Branchinectellinae

Branchinectella media (Schmankewitsch, 1873) 1 Bmed

Subfamily Chirocephalinae

Chirocephalus diaphanus Prevost, 1803 52 Cdia

Family Streptocephalidae

Streptocephalus torvicornis (Waga, 1842) 17 Stor

Family Tanymastigidae

Tanymastigites perrieri (Daday, 1910) 8 Tper

Family Thamnocephalidae

Phallocryptus spinosus (Milne-Edwards, 1840) 18 Pspi

Spinicaudata

Family Cyzicidae

Cyzicus tetracerus (Krynicki, 1830) 20 Ctet

Family Leptestheridae

Leptestheria mayeti (Simon, 1885) 8 Lmay

Notostraca

Family Triopsidae

Lepidurus apus lubbocki Brauer, 1873 14 Llub

Triops cancriformis (Bosc, 1801) 4 Tcan

Triops granarius (Lucas, 1864) 11 Tgra

Triops simplex Ghigi, 1921 22 Tsim

44 Hydrobiologia (2016) 782:37–51

123

up to 20 km); this pattern indicated an abrupt

change in species composition with increasing

geographical distance. Unlike species from northern

sites, gradual changes in species composition with

geographical distance were observed for species

inhabiting the southern ecological zones; in this

case, the first three distance classes showed signifi-

cant positive autocorrelation (i.e. up to about

100 km), while the last class showed a significant

negative autocorrelation (P\ 0.05).

Bsch

Tper

StorCdia

Bmed

Bfer

Pspi

Asal

Ctet

Lmay

Llub

Tgra

TcanTsim

SizeCond

Turb

Macr

Tmax

PrecAET

dbMEM1

dbMEM2

dbMEM5

dbMEM4

dbMEM3

dbMEM11

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

-0.6 -0.4 -0.2 0 0.2 0.4 0.6

rm d

bRD

A2

rm dbRDA1

b

SizeCond

Turb

Macr

Tmax

PrecAET

dbMEM1

dbMEM2

dbMEM5

dbMEM4

dbMEM3

dbMEM11

-60

-40

-20

0

20

40

60

-60 -40 -20 0 20 40 60

dbR

DA

2 (2

7.7%

-15

%)

dbRDA1 (52.3% - 28.4%)

a

NW Mediterranean area

E Mediterranean and Tell

Steppe and desert area

Fig. 4 Map of the results of

distance-based redundancy

analysis (dbRDA) using the

first two axes (explained

constrained variation 80%,

unconstrained 43.4%).

Vectors representing spatial

and environmental variables

are superimposed on the

maps of sites (a) and species(b) using their multiple

correlation coefficients (rm)

with canonical axes.

(a) Constrained and

unconstrained variation for

each axis reported in

brackets; pond areas are the

same as in Fig. 2.

(b) Species acronyms listed

in Table 1

Hydrobiologia (2016) 782:37–51 45

123

Discussion

Diversity across taxonomic groups and pond areas

The large branchiopod fauna of Tunisia is rather rich

and diverse, and mostly similar to that of the

neighbouring Algeria (Gauthier, 1928b; Samraoui

et al., 2006). Among them we report the presence of

Branchinectella media and Branchinecta ferox

(Milne-Edwards, 1840) several decades after their last

report for the country (Gauthier, 1928a; Massal, 1951,

but see also Marrone et al., 2016).

Four major ecological groups within the Tunisian

large branchiopods were observed: (i) species exclu-

sive or strictly linked to the steppic area of the country

(i.e. Tanymastigites perrieri (Daday, 1910), Triops

granarius, Leptestheria mayeti Simon, 1885, and

Branchipus schaefferi); (ii) species linked to the wet

Mediterranean area (i.e.Chirocephalus diaphanus and

Lepidurus apus lubbocki); (iii) halophilic species

linked with the heavily mineralised water of sebkhas

and chotts (i.e. Artemia salina, Phallocryptus spino-

sus, and Branchinectella media); (iv) more euryecious

species, including all the other taxa. The halophilic

species and their habitats are more common in the

steppic and desert areas, but they occur in lower

density all along the country, with the only exception

of the northern slope of the Medjerda mountains.

Quite a sharp latitudinal segregation is present in

the notostracans, with Lepidurus apus lubbocki

confined to the more humid areas of the country

and followed, along a north-to-south transect, by

Triops cancriformis (Bosc, 1801), T. simplex Ghigi,

1921 and T. granarius, paralleling their arrangement

along the first axis of dbRDA. Although the Triops

species distribution areas show a certain overlap,

their distribution barycentres are clearly separated:

T. granarius is confined to the Sahel, T. simplex is

spread north of the Sahel in Tell steppes and in the

driest areas of the Mediterranean part of the country,

and T. cancriformis confined to the Mediterranean

area, being replaced by Lepidurus apus lubbocki in

the more humid areas.

The floodplain of the Medjerda river is a hotspot of

large branchiopod diversity in the Mediterranean

region; its richness is comparable to that of the Chaouia

Plain in Morocco (Thiery, 1991): an area of few square

kilometres hosts 9 large branchiopod species (Marrone

et al., 2016), with large branchiopods multiple co-

existence of up to 7 species in a single site. It is

noteworthy that the Medjerda floodplain is the only

area of Tunisia where it is possible to observe the co-

existence of large branchiopod species typical of the

steppic and of the more humid areas of the country,

these two assemblages being usually strictly allopatric.

Medjerda floodplain high faunal diversity was also

17% 24% 15%Environmental Joint Spatial

44%Residual

Fig. 5 Variation partition using partial-dbRDAs (explained

variance 56%), showing the pure and shared influence of

ecological and spatial variables on large branchiopod species

composition; numbers represent the percentage of variation

explained by each component

Fig. 6 Mantel’s autocorrelogram of the species distribution

similarity matrix; sites closer than about 100 km are signifi-

cantly autocorrelated, while sites more than 100 km apart are

independent or negatively autocorrelated, and significantly

differ in species composition

46 Hydrobiologia (2016) 782:37–51

123

observed in other crustacean taxa (Marrone, unpub-

lished data) and amphibians (Sicilia et al., 2009).

Spatial and environmental drivers explaining

species distribution patterns

Despite the extensive distribution of several large

branchiopod species in Tunisia, there were distinct

differences across ecoclimatic zones, suggesting that

factors other than dispersal may influence species

distribution patterns. Three main environmental vari-

ables (precipitation, air temperature, electrical con-

ductivity) and, to a lesser extent, turbidity, macrophyte

cover, and annual evapotranspiration explain a signif-

icant amount (17%) of the variation in large bran-

chiopod distribution across Tunisia.

Precipitation and maximum air temperature are

strongly associated with a steep North–South eco-

climatic gradient from the Mediterranean coastline

(with high precipitation and lower temperatures in

the north-western part of the country) to the southern

arid areas. The strong relationship of large bran-

chiopod assemblages with climatic variables sug-

gests that this taxonomic group should be a sensitive

indicator of climate change at larger scales, sensi-

tivity which was not confirmed in smaller study

areas (Waterkeyn et al., 2009). Furthermore, precip-

itation and air temperature, together with actual

evapotranspiration (higher in the western Mediter-

ranean area), influence hydroperiod length, which is

a key factor in shaping Mediterranean temporary

pond communities (Gascon et al., 2005; Marrone

et al., 2006; Waterkeyn et al., 2008, 2009).

The influence of salinity, which is the main variable

driving the conductivity gradient in Tunisia, is a well-

known factor shaping large branchiopod assemblages

in northern Africa (Samraoui et al., 2006) and in the

whole Mediterranean area (Waterkeyn et al., 2008).

Salinity was considered the most important habitat

factor negatively influencing the distribution of fresh-

water large branchiopods, adversely affecting the

density and survival of hatchings (Waterkeyn et al.,

2009), while hosting peculiar brackish water large

branchiopod assemblages (Samraoui et al., 2006).

Finally, inorganic turbidity is an environmental

variable known to affect large branchiopod diversity

and distribution (Waterkeyn et al., 2009). The

increased surface area of clay particles for the

attachment of organic matter and bacteria, enhancing

diversity of food types and reducing the efficiency of

visual predators, could explain the influence of turbid

waters on branchiopod assemblages (Woodward &

Kiesecker, 1994). On the other hand, the high turbidity

in wetlands containing large branchiopods could also

be the result of bioturbation caused by the feeding

behaviour of Notostraca and Spinicaudata themselves

(Luzier & Summerfelt, 1997).

Precipitation, air temperature, conductivity, and, at a

smaller scale, turbidity and macrophyte cover (i.e.

habitat complexity) are spatially structured. This is

supported by our results showing that precipitation and

air temperature have an effect on species composition

together with the broad-scale dbMEM1, while conduc-

tivity (correlated with pond size, being several saline

lakes large saline wetlands) and turbidity are linked to

the medium-scale dbMEM5, and macrophyte cover to

the medium-scale dbMEM4. The other medium- to

small-scale distance-based Moran’s Eigenvector Maps

(dbMEM2, dbMEM3, and dbMEM11) do not appear to

be significantly correlated with the above-mentioned

variables. They are closely linked with some species

widely distributed in the Sahel steppic area, like

Branchipus schaefferi and Triops granarius, as clearly

demonstrated by dbRDA analysis, suggesting that

spatial factors (i.e. pond density and their geographical

distance in a given area) alone may be important in

structuring their distribution pattern.

Using variation partition, indeed, we found that both

the pure spatial and environmental components

accounted for a significant fraction of the variation in

species composition, but that the joint variation, due to

spatially structured environmental variables, explain

24% of total variation (i.e. about one half of the total

variation explained by dbRDA analysis). Spatial

factors alone explain 15% of total variation: without

neglecting the possibility that there might be an effect

of unmeasured, spatially structured environmental

factors, this part of variation may be due to pure

spatial processes, such as dispersal limitation (Lan-

deiro et al., 2011). These results point to a combination

of different processes explaining large branchiopod

distribution in Tunisia, in which both dispersal abilities

and environmental species sorting (i.e. environmental

filters that may prevent the successful colonization of a

new pond, see Incagnone et al., 2015) play key roles.

These filters are known to act at as a regional scale in

those areas where sharp climatic gradients exist

(Naselli-Flores & Padisak, 2016), like in Tunisia.

Hydrobiologia (2016) 782:37–51 47

123

Finally, unexplained variation is 44% of total variation,

and may be linked to describers not used in the present

study (like biotic interactions), or simply due to

random factors. This value is quite low if compared

to other studies of this kind performed at the regional

scale on a wide range of freshwater taxa both in ponds

(85-95%: De Bie et al., 2012) and in lakes (81–92%:

Beisner et al., 2006), confirming the high importance of

the selected spatial and environmental variables in

shaping species distribution at the regional scale in

Tunisia.

Recent evidence supports the idea that successful

dispersal in large branchiopods is usually lower than

that expected based on the presence of resting eggs,

which can be easily carried by biological and

physical vectors (see discussions in De Meester

et al., 2002; Orsini et al., 2013; Incagnone et al.,

2015). Our findings from the inspection of the

Moran correlograms show that spatial autocorrela-

tion of the species composition matrix is positive

only below an inter-sites distance of 100 km, this

distance being higher in steppic and desert areas than

in the northern, wetter Mediterranean area (where it

is settled at about 20 km). This is in good accor-

dance with the growing evidence that the realized

dispersal in large branchiopods is higher in arid

areas and open grassland than in areas characterized

by higher habitat complexity due to vegetation

structure surrounding the ponds (Boven et al.,

2008; Korn et al., 2010, and references therein).

Based on mtDNA sequence data, Ketmaier et al.

(2012) found non-random, positively autocorrelated

spatial genetic structure only for inter-populations

distances lower than 23 km in the ‘‘Mediterranean

plain forest-dwelling’’ anostracan Chirocephalus

kerkyrensis Pesta, 1936, a value in close agreement

with the distance calculated for the Mediterranean

area of Tunisia using Mantel correlograms. Con-

versely, Schwenter et al. (2012) registered the

existence of significant avian-mediated gene-flow

among Spinicaudata populations occurring over very

large distances in arid Australia ([1000 km, over an

area of 800,000 km2), showing that under certain

conditions long-distance passive dispersal events

might be frequent. Such a pattern is ascribed to the

higher probability of dispersal by waterbirds, the

most effective vectors for long-range passive dis-

persal (Green et al., 2008) in open areas, where

water bodies are more attractive to waterbirds than

those surrounded by forest or shrubs (Tourenq et al.,

2001; Korn et al., 2010; Ketmaier et al., 2012).

These results suggest that, in agreement with our

hypothesis, both dispersal limitation and species

response to spatially structured environmental gradi-

ents may be involved in determining large branchiopod

distribution in Tunisian temporary water bodies.

Maximum inter-ponds distance to ensure metacom-

munity dynamics is therefore lower in Mediterranean

Tunisia compared to the more arid parts of the country.

It implies that a decrease in pond number and density in

the more humid area of the country exerts an impact on

the conservation of temporary pond biological diver-

sity that is stronger than that which may occur in the

steppic areas. This is quite regrettable, since the most

humid areas are also those which are more subject to

human use. In particular, the floodplain areas where the

highest species richness and the richest large bran-

chiopod assemblages were observed (i.e. those occur-

ring in the Medjerda plain) are the most heavily

cultivated in Tunisia, and the wetlands that survived to

date are under threat by agriculture and pollution, as

already observed in some Moroccan plains (van den

Broeck et al., 2015). Considering the large bran-

chiopod keystone role in temporary aquatic habitats

(Waterkeyn et al., 2011), the impacts on these organ-

isms may have negative consequences on the structure

and survival of the entire pond communities and have to

be carefully considered on a regional basis when

conservation plans are developed.

Acknowledgments Alessandra Sicilia is gratefully

acknowledged for the support provided in the field work.

Sampling was partially granted by a ‘‘Borsa di perfezionamento

all’estero’’ of the University of Palermo (2005\2006) given to

FM.

References

Alonso, M., 1996. Crustacea, Branchiopoda. Fauna Iberica,

Museo Nacional de Ciencias Naturales. CSIC, Madrid.

Angeler, D. G., O. Viedmaa, S. Sanchez-Carrillo &M. Alvarez-

Cobelas, 2008. Conservation issues of temporary wetland

Branchiopoda (Anostraca, Notostraca: Crustacea) in a

semiarid agricultural landscape: What spatial scales are

relevant? Biological Conservation 141: 1224–1234.

Beisner, B. E., P. R. Peres-Neto, E. S. Lindstrom, A. Barnett &

M. L. Longhi, 2006. The role of environmental and spatial

processes in structuring lake communities from bacteria to

fish. Ecology 87: 2985–2991.

48 Hydrobiologia (2016) 782:37–51

123

Borcard, D. P., P. Legendre & P. Drapeau, 1992. Partialling out

the spatial component of ecological variation. Ecology 73:

1045–1055.

Borcard, D. & P. Legendre, 2012. Is the Mantel correlogram

powerful enough to be useful in ecological analysis? A

simulation study. Ecology 93: 1473–1481.

Boven, L., B. Vanschoenwinkel, E. R. M. De Roeck, A. Huls-

mans & L. Brendonck, 2008. Diversity and distribution of

large branchiopods in Kiskunsag (Hungary) in relation to

local habitat and spatial factors, Implications for their

conservation. Marine and Freshwater Research 59:

940–950.

Brendonck, L., D. C. Rogers, J. Olesen, S.Weeks &W. R. Hoeh,

2008. Global diversity of large branchiopods (Crustacea,

Branchiopoda) in freshwater. Hydrobiologia 595: 167–176.

Champeau, A. & A. Thiery, 1990. Les crustaces entomostraces

des eaux stagnantes de Corse. Importance particuliere des

especes monovoltines mediterraneennes de Copepodes

Calanoıdes et d’Anostraces dans le sud-est de l’ıle. Bulletin

de la Societe Zoologique de France 115(1): 55–75.

Colburn, E. A., 2004. Vernal Pools. Natural History and Con-

servation. McNaughton & Gunn, Inc., Saline.

Colwell, R. K., 2005. EstimateS: statistical estimation of species

richness and shared species from samples, version 9.1.

http://purl.oclc.org/estimates. Accessed 12 Oct 2015.

De Bie, T., L. De Meester, L. Brendonck, K. Martens, B.

Goddeeris, D. Ercken, H. Hampel, L. Denys, L. Vanhecke,

K. Van der Gucht, J. Van Wichelen, W. Vyverman & S.

A. J. Declerck, 2012. Body size and dispersal mode as key

traits determining metacommunity structure of aquatic

organisms. Ecology Letters 15: 740–747.

De Meester, L., A. Gomez, B. Okamura & K. Schwenk, 2002.

The Monopolization Hypothesis and the dispersal-gene

flow paradox in aquatic organisms. Acta Oecologica 23:

121–135.

Demeter, L. & A. Stoicescu, 2008. A review of the distribution

of large branchiopods (Branchiopoda: Anostraca, Noto-

straca, Spinicaudata, Laevicaudata) in Romania. North-

Western Journal of Zoology 4: 203–223.

Dray, S., 2009. packfor: forward Selection with Permutation

(Canoco p. 46). R package version 0.0-7/r58. http://R-

Forge.R-project.org/projects/sedar/.Accessed 12Oct 2015.

Dray, S., R. Pelissier, P. Couteron, M. J. Fortin, P. Legendre, P.

R. Peres-Neto, E. Bellier, R. Bivand, F. G. Blanchet, M. De

Caceres, A. B. Dufour, E. Heegaard, T. Jombart, F. Munoz,

J. Oksanen, J. Thioulouse & H. H. Wagner, 2012. Com-

munity ecology in the age of multivariate multiscale spatial

analysis. Ecological Monographs 82: 257–275.

Dumont, H. J. & S. Negrea, 2002. Introduction to the class

Branchiopoda. Guides to the Identification of the

Microinvertebrates of the ContinentalWaters of theWorld.

Backhuys Publishers, Kerkwerve.

Dumont, H. J., P. Laureys & J. Pensaert, 1979. Anostraca,

Conchostraca, Cladocera and Copepoda from Tunisia.

Hydrobiologia 66: 259–274.

Furrer, R., D. Nychka & S. Sain, 2011. Fields: tools for spatial

data. R package version 6.6. http://CRAN.R-project.org/

package=fields. Accessed 12 Oct 2015.

Gascon, S., D. Boix, J. Sala & X. D. Quintana, 2005. Variability

of benthic assemblages in relation to the hydrological

pattern in Mediterranean salt marshes (Emporda wetlands,

NE Iberian Peninsula). Archiv fur Hydrobiologie 163:

163–181.

Gaston, K. J. & J. I. Spicer, 2004. Biodiversity: An Introduction,

2nd ed. Wiley-Blackwell Ltd., Malden.

Gauthier, H., 1928a. Recherches Sur La Faune Des Eaux Con-

tinentales De L’algerie Et De La Tunisie. Imp. Minerva,

Alger: 416.

Gauthier, H., 1928b. Description du Tanymastix mzabica, nou-

veau Phyllopode Anostrace du Sahara et liste des Anos-

traces recoltes recemment par l’auteur en Algerie et en

Tunisie. Bulletin de la Societe d’Histoire Naturelle de

l’Afrique du Nord 19: 87–90.

Goslee, S. & D. Urban, 2013. ecodist: dissimilarity-based

functions for ecological analysis. R package version 1.2.9.

http://CRAN.R-project.org/package=ecodist. Accessed 12

Oct 2015.

Gotelli, N. J. & R. K. Colwell, 2001. Quantifying biodiversity:

procedures and pitfalls in the measurement and comparison

of species richness. Ecology Letters 4: 379–391.

Green, A. J., K. M. Jenkins, D. Bell, P. J. Morris & R.

T. Kingsford, 2008. The potential role of waterbirds in

dispersing invertebrates and plants in arid Australia.

Freshwater Biology 53: 380–392.

Gurney, R., 1909. On the freshwater Crustacea of Algeria and

Tunisia. Journal of the Royal Microscopical Society 443:

273–305.

Hijmans, R. J., S. E. Cameron, J. L. Parra, P. G. Jones & A.

Jarvis, 2005. Very high resolution interpolated climate

surfaces for global land areas. International Journal of

Climatology 25: 1965–1978.

Incagnone, G., F. Marrone, L. Robba, R. Barone & L. Naselli-

Flores, 2015. How do freshwater organisms cross the ‘‘dry

ocean’’? A review on passive dispersal and colonization

processes with a special focus on temporary ponds.

Hydrobiologia 750: 103–123.

Ketmaier, V., F. Marrone, G. Alfonso, K. Paulus, A. Wiemann,

R. Tiedemann & G. Mura, 2012. Mitochondrial DNA

regionalism and historical demography in the extant pop-

ulations of Chirocephalus kerkyrensis (Branchiopoda:

Anostraca). PLoS One 7(2): e30082.

King, J. L., 1998. Loss of diversity as a consequence of habitat

destruction in California vernal pools. InWitham, C.W., E.

Bauder, D. Belk, W. Ferren & R. Ornduff (eds), Ecology,

Conservation, and Management of Vernal Pool Ecosys-

tems: Proceedings from a 1996 Conference. California

Native Plant Society, Sacramento: 119–123.

Korn, M., F. Marrone, J. L. Perez-Bote, M. Machado, M. Cristo,

L. Cancela da Fonseca & A. K. Hundsdoerfer, 2006. Sister

species within the Triops cancriformis lineage (Crustacea,

Notostraca). Zoologica Scripta 35: 301–322.

Korn, M., A. J. Green, M. Machado, J. Garcıa-de-Lomas, M.

Cristo, L. Cancela da Fonseca, D. Frisch, J. L. Perez-Bote

& A. K. Hundsdoerfer, 2010. Phylogeny, molecular ecol-

ogy and taxonomy of southern Iberian lineages of Triops

mauritanicus (Crustacea: Notostraca). Organism Diversity

and Evolution 10: 409–440.

Korn, M., N. Rabet, H. V. Ghate, F. Marrone & A. K. Hunds-

doerfer, 2013. Molecular phylogeny of the Notostraca.

Molecular Phylogenetics and Evolution 69: 1159–1171.

Landeiro, V. L., W. E. Magnusson, A. S. Melo, H. M. V.

Espırito-Santo & L. M. Bini, 2011. Spatial eigenfunction

Hydrobiologia (2016) 782:37–51 49

123

analyses in stream networks: do watercourse and overland

distances produce different results? Freshwater Biology

56: 1184–1192.

Legendre, P. & M. J. Anderson, 1999. Distance-based redun-

dancy analysis: testing multispecies responses in multi-

factorial ecological experiments. Ecological Monographs

69: 1–24.

Legendre, P. & L. Legendre, 2012. Numerical Ecology, 3rd ed.

Elsevier, Oxford.

Legendre, P., D. Borcard, G. Blanchet & S. Dray, 2010. PCNM:

PCNM spatial eigenfunction and principal coordinate

analyses. R package version 2.1. http://R-Forge.R-project.

org/projects/sedar/. Accessed 12 Oct 2015.

Luzier, J. M. & R. C. Summerfelt, 1997. Experimental

demonstration of the effects of clam shrimp on turbidity of

microcosms. The Progressive Fish-Culturist 59: 68–70.

Marrone, F., R. Barone & L. Naselli-Flores, 2006. Ecological

characterization and cladocerans, calanoid copepods and

large branchiopods of temporary ponds in a Mediterranean

island (Sicily, southern Italy). Chemistry and Ecology

22(suppl. 1): 181–190.

Marrone, F., M. Korn, F. Stoch, L. Naselli-Flores & S. Turki,

2016. Updated checklist and distribution of large bran-

chiopods (Branchiopoda: Anostraca, Notostraca, Spini-

caudata) in Tunisia. Biogeographia (in press).

Massal, L., 1951. Note sur la presence en Tunisie de Bran-

chinecta ferox (M. Edw.) (Crustace Phyllopode Anos-

trace). Bulletin de la Societe des Sciences Naturelle de

Tunisie 4: 41–44.

Naselli-Flores, L. & R. Barone, 2012. Phytoplankton dynamics

in permanent and temporary Mediterranean waters: is the

game hard to play because of hydrological disturbance?

Hydrobiologia 698: 147–159.

Naselli-Flores, L. & J. Padisak, 2016. Blowing in the wind: how

many roads can a phytoplanktont walk down? A synthesis

on phytoplankton biogeography and spatial processes.

Hydrobiologia 764(1): 303–313.

Naselli-Flores, L., R. Termine & R. Barone, 2016. Phyto-

plankton colonization patterns. Is species richness

depending on distance among freshwaters and on their

connectivity? Hydrobiologia. doi:10.1007/s10750-015-

2283-4.

Nhiwatiwa, T., L. Brendonck, A. Waterkeyn & B. Vanschoen-

winkel, 2011. The importance of landscape and habitat

properties in explaining instantaneous and long-term dis-

tributions of large branchiopods in subtropical temporary

pans. Freshwater Biology 56: 1992–2008.

O’Brien, E. M., 2006. Biological relativity to water-energy

dynamics. Journal of Biogeography 33: 1868–1888.

Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, R.

B. O’Hara, G. L. Simpson, P. Solymos, M. H. H. Stevens &

H. Wagner, 2011. vegan: community ecology package. R

package version 1.17-12. http://CRAN.R-project.org/

package=vegan. Accessed 12 Oct 2015.

Orsini, L., J. Vanoverbeke, I. Swillen, J. Mergeay & L. De

Meester, 2013. Drivers of population genetic differentia-

tion in the wild: isolation by dispersal limitation, isolation

by adaptation and isolation by colonization. Molecular

Ecology 22: 5983–5999.

Peel, M. C., B. L. Finlayson & T. A. McMahon, 2007. Updated

world map of the Koppen-Geiger climate classification.

Hydrology and Earth System Sciences 11: 1633–1644.

Peres-Neto, P. R., P. Legendre, S. Dray & D. Borcard, 2006.

Variation partitioning of species data matrices: estimation

and comparison of fractions. Ecology 87: 2614–2625.

QGIS Development Team, 2015. QGIS geographic information

system. open source geospatial foundation. http://qgis.

osgeo.org. Accessed 12 Oct 2015.

R Core Team, 2015. R: A language and environment for statis-

tical computing. R Foundation for Statistical Computing,

Vienna. http://www.R-project.org. Accessed 12 Oct 2015.

Roux, P. & A. Thiery, 1988. Complement a la repartition des

Crustaces Branchiopodes Anostraca, Notostraca et Con-

chostraca du Maroc: le Maroc oriental. Bulletin de la

Societe d’Histoire Naturelle du Toulouse 124: 225–233.

Samraoui, B. & H. J. Dumont, 2002. The large branchiopods

(Anostraca, Notostraca and Spinicaudata) of Numidia

(Algeria). Hydrobiologia 486: 119–123.

Samraoui, B., K. Chakri & F. Samraoui, 2006. Large bran-

chiopods (Branchiopoda: Anostraca, Notostraca and

Spinicaudata) from the salt lakes of Algeria. Journal of

Limnology 65: 83–88.

Schwenter, M., B. V. Timms & S. Richter, 2012. Flying with the

birds? Recent large-area dispersal of four Australian Lim-

nadopsis species (Crustacea: Branchiopoda: Spinicaudata).

Ecology and Evolution 2: 1605–1626.

Scott, D. W., 2009. Sturge’s rule. Wiley Interdisciplinary

Reviews: Computational Statistics 1: 303–306.

Sicilia, A., F. Marrone, R. Sindaco, S. Turki & M. Arculeo,

2009. Contribution to the knowledge of Tunisian

amphibians: notes on distribution, habitat features and

breeding phenology. Herpetology Notes 2: 107–132.

Sokol, E. R., J. M. Hoch, E. Gaiser & J. C. Trexler, 2014.

Metacommunity structure along resource and disturbance

gradients in Everglades wetlands. Wetlands 34(suppl 1):

135–146.

Thiery, A., 1987. Les Crustaces Branchiopodes Anostraca,

Notostraca et Conchostraca des milieu limniques tempo-

raires (Dayas) au Maroc. Taxonomie, Biogeographie,

Ecologie. 405 pp. These Doctorat d’Etat es Sciences,

Universite Aix-Marseille III.

Thiery, A., 1991. Multispecies coexistence of Branchiopods

(Anostraca, Notostraca and Spinicaudata) in temporary

ponds of Chaouia plain (western Morocco): sympatry or

syntopy between usually allopatric species. Hydrobiologia

212: 117–137.

Thiery, A., 1996. Large branchiopods (Crustacea: Anostraca,

Notostraca, Spinicaudata, Laevicaudata) from temporary

inland waters of the Arabian Peninsula. Fauna of Saudi

Arabia 15: 37–98.

Tourenq, C., S. Aulagnier, L. Durieux, S. Lek, F. Mesleard, A.

Johnson & J. L. Martin, 2001. Identifying rice fields at risk

from damage by the Greater Flamingo. Journal of Applied

Ecology 38: 170–179.

Trabucco, A. & R. J. Zomer, 2010. Global soil water balance

geospatial database. CGIAR consortium for spatial information.

http://www.cgiar-csi.org/data/global-high-resolution-

soil-water-balance. Accessed 12 Oct 2015.

50 Hydrobiologia (2016) 782:37–51

123

Turki, S. & A. El Abed, 1999. Nouvelles informations sur les

Copepodes calanoıdes et cyclopoıdes des eaux continen-

tales tunisiennes. Crustaceana 72: 157–169.

Turki, S. & B. Turki, 2010. Copepoda and Branchiopoda from

Tunisian temporary waters. International Journal of Bio-

diversity and Conservation 2(4): 86–97.

Van den Broeck, M., A. Waterkeyn, L. Rhazi & L. Brendonck,

2015. Distribution, coexistence, and decline of Moroccan

large branchiopods. Journal of Crustacean Biology 35:

355–365.

Waterkeyn, A., P. Grillas, B. Vanschoenwinkel & L. Brendon-

ck, 2008. Invertebrate community patterns in Mediter-

ranean temporary wetlands along hydroperiod and salinity

gradients. Freshwater Biology 53: 1808–1822.

Waterkeyn, A., P. Grillas, E. R. M. De Roeck, L. Boven & L.

Brendonck, 2009. Assemblage structure and dynamics of

large branchiopods in Mediterranean temporary wetlands:

patterns and processes. Freshwater Biology 54: 1256–1270.

Waterkeyn, A., P. Grillas, M. Anton-Pardo, B. Vanschoen-

winkel & L. Brendonck, 2011. Can large branchiopods

shape zooplankton communities in Mediterranean tempo-

rary wetlands?Marine and Freshwater Research 62: 46–53.

Williams, D. D., 2006. The Biology of Temporary Waters.

Oxford University Press, Oxford.

Woodward, B. D. & J. Kiesecker, 1994. Ecological conditions

and the notonectid—fairy shrimp interaction. The South-

western Naturalist 39: 160–164.

Zacharias, I. & M. Zamparas, 2010. Mediterranean temporary

ponds. A disappearing ecosystem. Biodiversity and Con-

servation 19: 3827–3834.

Hydrobiologia (2016) 782:37–51 51

123