Taxonomic diversity and toxicological assessment of Cyanobacteria in Moroccan inland waters Diversité taxonomique et évaluation toxicologique des cyanobactéries dans les eaux continentales du Maroc.

Abstract

Research on the ecology, biodiversity and toxicology of cyanobacteria in Moroccan inland waters has been carried out since 1994. The results demonstrate the existence of several taxa of cyanobacteria. Most of them are toxic, bloom‑forming species present in various water bodies of the country. The present study follows upon this earlier work and spans the 2003-2006 period. The major aim was to update and supplement the existing national cyanobacteria inventory and to isolate new toxic strains. During the study period, more than 40 aquatic environments were visited and sampled.

Almost 300 taxa of cyanobacteria were recorded. They belonged to 3 orders, 14 families and 46 genera. Among these, about 78 taxa are recorded for the first time in Morocco; 29 strains of cyanobacteria were successfully isolated and cultured in the laboratory. All the collected cyanobacteria, including natural blooms, mats, and cultured strains, were analyzed for toxicity and hepatotoxins (microcystins) were quantified. Using the High-performance liquid chromatography technique coupled to photodiode array (PDA) detector (HPLC-PDA), four samples of Microcystis blooms showed the presence of microcystins (MCs), with a concentration ranging between 1.87 and 64.4 µg•g^‑1 MC‑LR eq (microcystin-LR equivalents). A total of five different structural variants of MCs were detected (MC-LR, -RR, -YR, -FR, -WR). Furthermore, 3 of 29 isolates were confirmed as MCs producing strains.

The results show that the widening of the survey led to a better knowledge of the diversity of cyanobacteria. The taxonomic inventory was greatly increased and several cyanobacteria strains were characterized for their toxicity. The results should be useful as a database for the identification of various aquatic environments contaminated by cyanobacterial toxins (microcystins), which represent a potent sanitary risk for human and animals.

Key words:

• Cyanobacteria,
• taxonomy,
• microcystins,
• blooms,
• isolates,
• inland waters,
• Morocco

Résume

Au Maroc, les recherches sur l’écologie, la biodiversité et la toxicologie des cyanobactéries des eaux continentales ont été entamées à partir de 1994. Les résultats obtenus ont montré l’existence de plusieurs taxons de cyanobactéries, dont certains sont responsables de la formation de blooms toxiques. Faisant suite aux précédents travaux, la présente étude, réalisée lors de la période 2003-2006, où plus de 40 milieux aquatiques ont été prospectés, a pour objectif de compléter et d’actualiser l’inventaire national des cyanobactéries d’eau douce du Maroc et d’isoler de nouvelles souches toxiques.

Plus de 300 taxons de cyanobactéries appartenant à 3 ordres, 14 familles et 46 genres ont été inventoriés. À notre connaissance, 78 taxons sont cités pour la première fois au Maroc et 29 souches de cyanobactéries ont pu être isolées et cultivées en laboratoire. Le matériel cyanobactérien planctonique ou benthique collecté sur le terrain (blooms, écumes, films benthiques, etc.) et la biomasse des souches isolées produite en culture au laboratoire, ont été analysés pour l’évaluation de la toxicité et la quantification des cyanotoxines (microcystines).

L’utilisation de la technique HPLC-PDA (High-performance liquid chromatography technique coupled to photodiode array (PDA) detector) a permis d’identifier quatre blooms toxiques à Microcystis et la détection de microcystines (MCs) à des concentrations variant entre 1,87 et 64,4 µg•g^‑1 eq MC-LR (microcystin-LR equivalents). Cinq variantes structurales de MCs ont pu être détectées (MC-LR, -RR, -YR, -FR, -WR). Parmi les 29 souches isolées et produites au laboratoire, trois seulement ont confirmé la production de microcystines.

Les résultats obtenus constituent un apport substantiel à la taxonomie des cyanobactéries et à l’évaluation de la biodiversité des cyanobactéries du Maroc. Ces données peuvent être utilisées comme base pour identifier les milieux aquatiques potentiellement contaminés par les cyanobactéries et capables de générer un haut risque sanitaire pour les hommes et les animaux.

Mots clés:

• Cyanobactéries,
• taxonomie,
• microcystines,
• blooms,
• souches isolées,
• eaux continentales,
• Maroc

1. Introduction

Cyanobacteria (Prokaryotic blue green algae or cyanoprokaryotes) occur especially in freshwater and brackish environments, but may be found also in marine and terrestrial ecosystems. They may dominate the phytoplankton of lakes and rivers, forming respectively blooms and mats, when environmental conditions are appropriate for their growth (OLIVER et GANF, 2000).

Cyanobacteria have received growing attention as producers of a diverse array of biologically active natural products. Some of these compounds have potential biotechnological uses (THAJUDDIN et SUBRAMANIAN, 2005) and others, called “cyanotoxins”, appear to be toxic and pose a hazard to many organisms, including humans, who contact, ingest or use water contaminated by theses toxins (CODD et al., 2005, a review). Cyanotoxins include potent hepatotoxins, neurotoxins, cytotoxins, irritants or gastrointestinal toxins (WIEGAND et PFLUGMACHER, 2005, a review). Among hepatotoxins, the microcystins (MCs) are the best documented. Until now more than 75 natural MCs structural variants have been characterized (CODD et al., 2005, a review).

Over 100 species of cyanobacteria belonging to 40 genera are reported to be toxic (JAYATISSA et al., 2006). Microcystis is the most frequent bloom-forming cosmopolitan genus (CARMICHAEL, 1996).

The problem of massive proliferation of cyanobacteria seems more accentuated in temperate zones, known by a remarkable diversity. In these areas, cyanobacterial blooms are most prominent during late summer and early autumn. In Mediterranean regions and subtropical areas, the bloom season may start earlier and persist longer (COOK et al., 2004).

Benefiting from its subtropical geographical situation, Mediterranean climate and two maritime frontages, Morocco has a mosaic of naturally diversified aquatic ecosystems, in addition to more than 100 reservoirs. This great heterogeneity of habitats and biotopes offers varied ecological conditions leading to a rich and diversified micro-flora. Moreover, previous works, although relatively limited in time and space, revealed the presence of an un‑negligible cyanobacterial microflora. Since then, the national taxonomic list of cyanobacteria has increased, following the appearance of new taxonomic records in recent limnological studies (LOUDIKI et al., 2002).

Cyanobacteria blooms are common in some freshwater bodies used for recreation and/or as drinking water reservoirs. This is likely the result of a rapid eutrophication due to high nutrient loads and semi-arid climate factors (LOUDIKI et al., 2002 ; OUDRA et al., 2001a).

In Morocco, it is only in 1994 that a research program on diversity of cyanobacteria was started when many cyanobacteria bloom-forming species have been identified and studied. These results confirmed that Morocco, like other countries in the world, was affected by massive proliferation of toxic cyanobacteria (LOUDIKI et al., 2002 ; OUDRA et al., 2001a ; SABOUR et al., 2002, 2005 ; SBIYYAA et al., 1998).

This work, especially in its taxonomic part, represents a substantial addition to the North African and Mediterranean previous studies, which remain often limited (ABDEL‑RAHMAN et al., 1993 ; ABOAL et PUIG, 2005 ; ALBAY et al., 2003 ; BOUAÏCHA et NASRI, 2004 ; COOK et al., 2004 ; OUDRA et al., 2001a ; PRATI et al., 2002 ; SABOUR et al., 2002 ; VASCONCELOS et al., 2001).

The Moroccan taxonomic inventory of cyanobacteria remains relatively limited. Our study, which spans over the period between March 2003 and September 2006, aims firstly to extend the investigation to other and new aquatic environments and to update and supplement the existing cyanobacteria inventory, secondly, to isolate new cyanobacterial strains and thirdly, to assess their potent toxicity.

2. Materials and methods

2.1 Sites description and sampling

More than 40 various aquatic environments (10 natural lakes, 19 reservoirs, 3 lagoons and estuaries, 15 streams and rivers, etc.) including fresh and salted waters have been sampled. The choice of sites was based essentially on their trophic conditions and ecological importance. Some sites have biological and ecological interest, called SIBES (as Sebkha Zima, Dayèt Aoua, Assif Ait Mizane, Massa estuary, etc.) or are Ramsar sites (as Daya Sidi Boughaba, Al Massira reservoir, etc.). Other sites are used for drinking water production reservoirs (as Al Massira, Mansour Eddahbi, Youssef ben Tachfine, Sidi Mohamed Ben Abdellah, etc.) and /or are agricultural and recreational waters (as Lalla Takerkoust reservoir, etc.). These sites are located in different Moroccan hydrographic basins (32 00 N, 5 00 W), under various climatic conditions, at various altitudes and characterized by different trophic status. The sampling was done between 2003 and 2006 each time, at the summer and early autumn period, corresponding to the phases of massive cyanobacteria proliferation blooms and generally to the peak of cyanobacterial biomass.

In reservoirs and lakes, the sampling of planktonic cyanobacteria was done using a 37 µm mesh phytoplankton net. In the running waters (streams, rivers, channels, etc.), the sampling was carried out either using a 37 µm mesh net with handle or by scraping benthic species attached to various substrates (blocks, stones, sediments, vegetation, etc.).

The cyanobacteria bloom‑forming were collected in three reservoirs: Lalla Takerkoust (TAK), Al Massira (ALM), Mansour Eddahbi (MED) and from two natural lakes: Dayèt Erroumi (DE) and Tigalmamines (TI). Whereas, the cyanobacteria mat‑forming were collected in two streams: Oued Ourika (OUR) and Oued Oukaïmeden (OK). For some localities like TAK, OUR and OK, the sampling was realized several times according to the cyanobacteria bloom occurrences. In order to supplement the laboratory culture collection with more cyanobacteria strains, 29 isolates, planktonic or benthic, coming from 24 different water bodies, were isolated and cultured in laboratory conditions.

2.2 Taxonomic study

The taxonomic identification of the species was carried out under the light microscope. Species were observed, measured, and identified using several specialized cyanobacteria taxonomic books (KOMÁREK et ANAGNOSTIDIS 1999, 2005 ; STARMACH, 1966). Identification was based on morphologic criteria such as form of cells (width, length), filaments or colonies, their mucilaginous envelopes, coloration, pigmentation and mode of cell division.

2.3 Cyanobacteria isolation and strain culture conditions

Z8 medium (KOTAI, 1972) was used for the isolation and culture of cyanobacteria. Cultures were maintained at 25 °C ± 2 °C, at a light intensity of 82•µE•m^‑2•S^‑1 fluorescent continuous light and with a light/dark cycle of 16h/8h. The cyanobacterial biomass produced was harvested at the end of the exponential growth phase by flotation, centrifugation or filtration through a 37 µm pore net filter. The concentrated material was freeze‑dried and stored at -20 °C prior to toxicity assessment and MCs detection and quantification.

2.4 Toxicity assessment, detection and quantification of hepatotoxins (microcystins)

2.4.1 Mice bioassay

Cyanobacteria mammals’ toxicity was measured by intra-peritoneal (i.p) mouse bioassay, using 18-22 g male white Swiss mice. The method was already described by OUDRA et al. (2001b). Briefly, a determined mass of lyophilized cyanobacteria material was suspended in 0,9% NaCl. The obtained physiological solution was injected into pairs of mice per each dose level. Animal behaviour, toxicity symptoms and survival time were registered. After death, animals were autopsied internally for signs of hepatotoxicity (hemorrhagic shock, liver swelling, etc.). The toxicity was evaluated as LD₅₀ (mg•kg^‑1 dry weight) determination (LD₅₀ : Lethal dose for 50% mortality of tested animals). By the mice bioassay, the toxicity of three kinds of samples was evaluated: three natural cyanobacteria bloom-forming (TAK, ALM, MED), three natural cyanobacetria mat-forming (Nos. M., Lyn. A1, Lyn. A2) and two isolated cyanobacteria strains laboratory cultures (Pseud. C. S24., Pseud. G. S25).

2.4.2 Artemia salina bioassay

Cyanobacteria Crustacean toxicity was determined and evaluated by Artemia salina test only for four kinds of cyanobacterial samples, including natural blooms and strain cultures (Mic. A. TAK, Mic. A. MED, Mic. A. ALM, Mic. W. Ti). The method was already described by SABOUR et al. (2005). Briefly, the dried brine shrimp eggs were purchased from special stores. The larvae (8-20 in, each well of a multi well microtiter plate) were used in the tests 24 h after hatching. Toxic fractions were pre-purified from 75% methanol extracts by using an activated Octadecyl Silane - ODS gel cartridge. The toxicity of cyanobacterial fractions to Artemia salina larvae was tested in natural seawater, in loosely covered micro-plates at 25 °C ± 2 °C. The test results were expressed as percentage of dead and 24 h‑LC⁵⁰ (lethal concentration (µg dry weight•mL^‑1), for 50% of the tested animals), calculated by EPA Probit Analysis Program (Version 1.5).

2.4.3 Cyanotoxins (MCs) extraction

All collected cyanobacteria materials including blooms, mats and laboratory cultures, were subjected to hepatotoxins analysis (detection of hepta-peptide microcystins). Lyophilized biomass was extracted with 70% aqueous methanol (2 mg DW•mL^‑1) and dried by rotary evaporation at 45 ºC. Two extract types were obtained as follows: 1) a concentrated extract, by extracting twice the biomass with 150 µL of 70% methanol and 2) a dilute extract, by extracting with 500 µL of 70% methanol, the residue of the biomass extracted with the said 300 µL. This procedure ensures total recovery of MCs. All extracts were filtered through a 1.2 µm GF/C glass filter before being subjected to HP a photodiode array detector.

2.4.4 Microcystins HPLC-PDA analysis

Chromatographic analysis was performed by HPLC Waters equipment (model 2695) with a photodiode array detector (model 996). The column used was Chromolith C18 (250 mm x 4.6 mm, 5µm.). The mobile phase system was: a) H²O + 0.05% (v/v) trifluoroacetic acid (TFA) and b) Acetonitrile (MeCN) + 0.05% (v/v) TFA. A linear gradient of 30 to 70% for a) and 30 to 100% for b). The volume injected was 100 µL or 50 µL, as advisable, and the mobile phase run at 1 mL•min^‑1. MCs were identified on the basis of their UV‑spectra at 238 nm and retention time. Standard MC-LR, -YR, -RR, -LF and -LW were purchased from Calbiochem (Germany). MC-FR and -WR were purified in the Autonomous University of Madrid laboratory. Other MCs different from the ones said earlier were quantified using microcystin-LR as a standard. All the chemicals used were of chromatographic grade (Scharlau Chemie Barcelona, Spain).

3. Results

3.1 Cyanobacterial biodiversity

Following the examination of the various existing cyanobacteria records in Moroccan waterbodies, the resulting inventory counted 377 taxa (DOUMA et al., 2004). This analysis allowed the extension of the list quoted by LOUDIKI et al. (2002), in which the total number of taxa inventoried at the time was 150. The additional survey allowed the inventory of 299 taxa, of which 78 species (26% of the total taxa) were observed for the first time in Morocco and belonged to 31 different genera. 54% of inventoried new taxa are among the 8 dominating genus which are: Phormidium, Lyngbya, Leptolyngbya, Pseudanabaena, Geitlerinema, Gloecapsopsis, Oscillatoria, Synechocystis (Table 1).

Table 1

New taxa inventoried for the first time in Moroccan inland waters.

Nouveaux taxons inventoriés pour la première fois dans les eaux continentales du Maroc.

Achroonema macromeres Sku. Anabaena oscillarioides Bory. Aphanocapsa conferta (W. et G.S. We.) Leg.et Cro. Aphanocapsa hyalina (Lyn.) Han. Aphanothece smithii Len. et Cron. Arthrospira maxima Set. et Gar. Chroococcopsis gigantea Geit. Chroococcus obliteratus Rich. Coelosphaerium kutzingianum Näg. Cyanobacterium minervae (Cop.) Kom. Cyanonephron styloides Hick. Cyanothece aeruginosa (Näg.) Kom. Geitlerinema acus (Cop.) An. Geitlerinema bigranulatum Tiw. Geitlerinema lemmermannii (Wol.) An. Gloecapsopsis dvorakii (Nov.) Kom. et An. Gloecapsopsis pleurocapsoides (Nov.) Kom. et An. Gloeocapsopsis magma (Bréb.) Kom. et An. Homoeothrix juliana (Born. et Flau.) Kir. Homoeothrix margalefii Kom. et Kal. Hormoscilla spongeliae Gom. Hormoscilla xishaensis Hu. Jaaginema homogeneum (Frém.) An. et Kom. Jaaginema unigranulatum (Bis.) An. Leptolyngbya faveolarum (Rab.) An. et Kom. Leptolyngbya benthonica (Sku.) An. Leptolyngbya boryana An. et Kom.

Leptolyngbya ercegovici (Cado) An. et Kom. Leptolyngbya fragilis (Gom.) An. et Kom. Leptolyngbya lurida (Gom.) An. et Kom. Limnothrix guttulata (Van Goor) Ume. et M. Wat. Limnothrix obliqueacuminata (Skuj.) Mef. Limnothrix pseudovacuolata (Uter.) An. Limnothrix rosea (Uter.) Meff. Lyngbya biebliana Clau. Lyngbya confervoides C. Ag. Lyngbya nigra Ag. Lyngbya semiplena J. Ag. Lyngbya stagnina Kütz. Nostoc Punctiforme (Kütz.) Har. Oscillatoria acuminata Geit. et Rut. Oscillatoria nigra Ag. et Kut. Oscillatoria ornata Kütz. Pascherinema moniliforme (Pas.) De Ton. Phormidiochaete balearica (Born.e Flau.)Kom. Phormidiochaete nordstedtii Kom. et Kan. Phormidium alaskense Gom. Phormidium bekesiense (I. Kiss) K. Kiss Phormidium bohneri Schm. Phormidium bulgaricum (Kom.) Kom. et An.  Phormidium chalybeum ( Mer.) An. et Kom. Phormidium chungii Gar. Phormidium fonticolum Kütz. Phormidium kolkwitzii Kom.

Phormidium lacustre (Cad.) An. Phormidium lusitanicum Samp. Phormidium paulsenianum (Boy.) Nov. Phormidium rosea (Skuj.) An. Phormidium simplicissimum (Gom.)An.et Kom. Phormidium stagninum (Kütz. ex Gom.) An. Phormidium tergestinum An. Phormidium versicolor Wart. Pseudanabaena biceps Bour. Pseudanabaena lonchoides An. Pseudanabaena moniliformis Kom. Pseudanabaena papillaterminata Kuk. Pseudanabaena starmachii An. Pseudocapsa sphaerica (Pro.) Kov. Romeria chlorina Böch. Schizothrix calcic la Gom. Spirulina nordstedtii Gom. Symploca cavernarum Cop. Synechococcus sciophilus sku. Synechocystis crassa Vorn. Synechocystis endobiotica Elen. et Holl. Synechocystis minuscule Vorn. Synechocystis pevalkii Erc. Woronichinia naegeliana (Ung.) Elen.

For many taxa (123), the taxonomic identification was stopped at the generic level. These taxa were also taken into account in the total national inventory, which currently counts 578 taxa (Figure 1).

Figure 1

Specific richness evolution of cyanobacteria in Moroccan inland waters.

Évolution de la richesse spécifique des cyanobactéries dans les eaux intérieures du Maroc.

The taxa listed in this study (299) were distributed in 3 orders, 14 families and 46 genera. Oscillatoriales constituted the more diversified order with 73% of the totality of taxa, followed by Chroococcales (17%) and Nostocales (10%) (Figure 2).

Figure 2

Spectrum of the orders of cyanobacteria inventoried in Moroccan inland waters.

Répartition des ordres de cyanobactéries dans les eaux intérieures marocaines.

This study enabled us to describe wild strains considered toxic or potentially toxic in literature reports. However, other strains were found to be toxic for the first time in our work.

Table 2 gives the geographical distribution of all these toxic strains in the Moroccan territory. It appears that Microcystis is the predominant phytoplanctonic bloom-forming genus. We present here a mapping of its distribution in some waterbodies (Figure 3). In addition, table 2 reveals the presence of two filamentous species (Nostoc muscorum and Lyngbya attenuata), which are responsible for mat‑forming in some particular aquatic ecosystems (particularly in stream ecosystems).

Table 2

Distribution of toxic and potentially toxic cyanobacterial strains in some Moroccan inland waters that were surveyed.

Distribution des souches toxiques et potentiellement toxiques dans certains milieux hydriques continentaux marocains prospectés.

SPECIES	RESERVOIRS
	AB	AM	EK	IB	IM	KT	LT	ME	NAK	OE	SM	SMBA	YBT
Anabaena variabilis Kütz.		*	*		*							*	*
Coelosphaerium naegelianum Ung.		*					*					*
Gomphosphaeria aponina Kütz.		*										*
Lyngbya birgei G. M. Smit.	*								*
Microcystis aeruginosa (Kütz.) Kütz.	*	*	*	*	*		*			*	*	*	*
Microcystis flos-aquae (wit.) Kirch.	*				*	*	*	*				*
Phormidium formosum (Bor.) An. et Kom.			*
Planktothrix agardhii (Gom.) An. et Kom.	*		*					*		*
Planktothrix isothrix (Skuj.) Kom. et Koma.	*	*									*		*
Pseudanabaena catenata Laut.							*		*
Pseudanabaena galeata Böch.							*
Pseudanabaena mucicola Nau. et Hub.		*
Schizothrix calcicola Gom.		*
Woronichinia naegeliana (Ung.) Elen
	NATURAL LAKES
	AA	AF	AW	DA	DE	IF	IFR	SBG	SZ	TI
Anabaena spiroides Kleb.									*
Anabaena variabilis Kütz.								*
Microcystis aeruginosa (Kütz.) Kütz.	*	*		*	*	*	*	*	*
Microcystis flos-aquae (wit.) Kirch			*
Microcystis wesenbergii (Kom.) Kom.										*
Phormidium formosum (Bor.) An. et Kom.								*	*
Pseudanabaena galeata Böch.									*
	RIVERS AND ESTUARIES
	AMI	BR	MAS	OK	OTE	OUR
Anabaena variabilis Kütz.		*	*
Nostoc muscorum Ag.	*			*		*
Phormidium formosum (Bor.) An. et Kom.					*
Pseudanabaena biceps Bour.						*

Reservoirs :	AB : Abdelmoumen ; AM : Al Massira ; EK : El Kansera ; IB : Ibn Batouta ; IM : Imfout ; KT : Kasba Tadla ; LT : Lalla Takerkoust ; ME : Mansour Eddahbi ; NAK : Nakhla ; OE : Oued El Makhazine ; SM : Smir ; SMBA : Sidi Mohamed Ben Abdellah ;YBT : Youssef Ben Tachfine.
Natural lakes :	AA : Aguelmam Azigza ; AF : Aguelmam Afourgah ; AW : Aguelmam Wiwane ; DA: Dayet Awa ; DE : Dayet Erroumi ; IF : Dayet Iffer ; IFR : Dayet Ifrah ; SBG: Merja de Sidi Boughaba ; SZ : Sebkha Zima ; TI : Tiguelmamines.
Rivers and estuaries :	AMI : Oued Aït Mizane ; BR : Bou Regreg estuary ; MAS : Estuaire Massa ; OK : Oued Oukaimden ; OTE: Oued Tensift OUR: Oued Ourika

Figure 3

Geographic localisation of some Moroccan reservoirs, natural lakes and ponds where Microcystis strains have been inventoried.

Localisation géographique des réservoirs, des lacs naturels et des étangs où des souches de Microcystis ont été inventoriées.

Reservoirs:

1- Ibn Battouta, 2- Smir, 3- M.B.A. Khettabi, 4- O. El Makhazine, 5- Sahela, 6- Mohammed V, 7- Idriss I, 8- Allal El Fassi, 9- Elkansera, 18- S. M. Ben Abdellah, 19- Oued Mellah, 20- Al Massira, 21- Imfout, 22- Daourat, 23- Lalla Takerkoust, 24- M. Eddahbi, 25- Abdelmoumen, 26- Y.B. Tachfine.

Natural lakes and ponds:

10- Sidi Boughaba, 11- Dayèt Erroumi, 12- Dayèt Aoua, 13-Dayèt Affourgah, 14- Dayèt Iffrah, Dayèt Aoua, Dayèt Iffrah, 15- Aguelmane Azigza, 16- Aguelmane Sidi Ali, 17- Tiguelmamines.

The diversity of the cyanobacteria is especially remarkable in the reservoirs (47% of the total specific richness), whereas, in natural lakes and the plain rivers, the biodiversity is respectively about 26% and 17% (Figure 4).

Figure 4

Distribution of inventoried cyanobacteria by type of aquatic biotope.

Distribution des cyanobactéries inventoriées selon les différents types de biotopes aquatiques.

3.2 Toxicological assessment and microcystins detection

The cyanobacteria taxonomic study was supplemented by a toxicological assessment leading to approve the toxicity of the collected cyanobacteria material both by mice and Artemia bioassays. Forty-two samples, among which 29 isolates, planktonic or benthic, coming from 24 different waterbodies, were analyzed for toxicity and/or detection of cyanobacterial hepatotoxins microcystins (MCs) (Table 3).

Table 3

Toxicological data of tested cyanobacteria (isolates, mats, blooms) in various Moroccan inland waters.

Données toxicologiques des cyanobactéries testées (isolats, blooms, films benthiques) dans divers milieux hydriques continentaux marocains.

	Sampling Date	Locality		Toxicity	HPLC Species
				Mouse bioassay LD50 (mg DW•kg‑1)	Artemia assay LC 50•24 h‑1 (μg•mL‑1)	Content (μg•g‑1 eq MC‑LR)	Microcystins profile ratio
Blooms	07‑2003	L.Takerkoust	Mic. A. TAK	176,8	1 414	13,94	RR (14), LR (16,6), FR (23,1) WR (46,4)
	10‑2004	M. Eddahbi	Mic. A. MED	352	1 712	64,4	RR (41,3),YR (21,7), LR (33,7), FR (1,1), WR (2,3)
	07‑2004	Al Massira	Mic. A. ALM	829	4 343	9,9	LR (100)
	10‑2004	Dayèt Erroumi	Mic. A. ER	-	-	1,87	RR (15,3) XZ* (8,2) LR (3,6) FR (20,3) WR (52,6)
Mats	10‑2004	Tiguelmamines	Mic. W. Ti	N.D.	N.D.	N.D.	N.D
	06‑2003	Oued Ourika	Nos. M. OUR	119,5	-	N.D.	N.D
	06‑2003	Oued Ourika	Lyn. A1	742	-	N.D.	N.D
	06‑2003	Oued‑Oukaimeden	Lyn. A2	1 450	-	N.D.	N.D
Isolates	07‑2003	L.Takerkoust	Mic. A. S13	-	-	9,44	RR (62), YR (2,6), LR (6,7) FR (7,9), WR (20,8)
	08‑2005	M. Eddahbi	Plank. A. S22	-	-	3,19	RR (100)
	12‑2004	Oued Ourika	Pseud. B. S23	-	-	0,28	XZ* (100)
	07‑2003	L.Takerkoust	Pseud. C. S24	1 192,7	-	N.D.	N.D
	06‑2003	Lac Zima	Pseud. G. S25	292,8	-	N.D.	N.D

* : Unknown MC variant, N.D. : Not detected

Plank. A. S22: Planktothrix agardhii; Pseud. B. S23: Pseudanabaena biceps ; Pseud. C. S.24: Pseudanabaena catenata; Pseud. G. S.25: Pseudanabaena galeata ; Mic. A. Tak, ALM, ER : Microcystis aeruginosa; Lyn. A1,2 : Lyngbya attenuata ; Nos. M. OUR : Nostoc Muscorum. Mic. W. Ti : Microcystis wesembergii. TAK: L. Takerkoust; ALM : Al Massira; MED : Mensou Eddahbi: OUR: Oued Ourika; OK: Oued Oukaimeden; LZ : Lac Zima; TI: Tiguelmamine.

3.2.1 Cyanobacteria “in situ” forming natural blooms and mats

As it was well indicated in table 3, the toxicity determined as LD⁵⁰ values for the Microcystis phytoplanktonic cyanobacteria blooms (Mic. A. TAK, Mic. A. MED and Mic. A. ALM) were respectively 176.8 mg•kg^‑1 DW, 352 mg•kg^‑1 DW and 829 mg•kg^‑1 DW. However, the mat-forming cyanobacterial, OUR mats (Nos. M., Lyn. A1) and OK mat (Lyn. A2) were respectively 119.5 mg•kg^‑1 DW, 742 mg•kg^‑1 DW and 1,450 mg•kg^‑1 DW (Table 3). During mice bioassays, both for the Microcystis forming blooms and Nostoc or Lyngbya forming mats, the survival time (1 to 4 h) and poisoning signs are similar to those observed for Microcystins hepatotoxicity.

For the three Microcystis natural blooms occurring at TAK, ALM and MED reservoirs, the toxicity was also evaluated by 24 h LC⁵⁰ Artemia biotest. The obtained values of 24 h LC⁵⁰ were respectively, 1.41 mg•mL^-1 , 1.71 mg•mL^‑1 and 4.34 mg•mL^-1 for the TAK, MED and ALM blooms.

For these collected natural bloom samples, the HPLC analysis revealed a clear difference of both for MCs content and variants composition, with a concentration of total MCs ranging between 1.87 and 64.4 µg•g^‑1 eq MC-LR (Table 3).

3.2.2 Cyanobacteria isolates strains (culture)

Twenty‑nine cyanobacteria strains belonging to the genera Phormidium, Pseudanabaena, Leptolyngbya, Microcystis, Synechocystis, Planktothrix, Oscillatoria, Lyngbya, Limnothrix, Jaaginema, Geitlerinema, Cyanobacterium and Anabaenopsis were isolated from natural samples collected from various waterbodies (reservoirs, lakes, rivers, streams, estuaries, irrigation channels, basins) (Table 4). Twelve strains were planktonic collected in the open water, forming or not scum or bloom. Seventeen are benthic collected from substrate such as gravel, mud in rivers, on the banks of reservoirs or attached to other filamentous species (Table 4).

Table 4

Isolated cyanobacteria strains screened for microcystin detection, sampling sites, type of biotope and sample characteristics.

«Screening» des souches isolées : détection de microcystins, sites échantillonnés, type de biotope, caractéristiques d’échantillons.

Strains code	Species	Sampling site / type of biotope / date	Characteristic of the strain in original sample
S1	Anabaena variabilis Kütz.	Bou Regreg / Estuary / 07‑2004	In the water mass, non forming bloom
S2	Anabanopsis circularis Gayr.	Sidi Bou Rhaba / Estuary / 10‑2004	In the water mass, non forming bloom
S3	Cyanobacterium minervae (Cop.) Kom	Palmeraie / Bassin, Marrakech / 06‑2003	In the water mass
S4	Geitlerinema lemmermannii.Tav.	Benslimane / Daya ( temporary lac) / 10‑2004	Attached to ground submerged or not by water
S5	Jaaginema sp.	Azigza / Lac / 11‑2004	Attached to the substrate of the banks or with other filamentous species
S6	Leptolyngbya faveolarum (Rab.) An. et Kom.	Sidi Bou Rhaba / Lac / 07‑2004	Attached to the substrate of the banks reservoir or with other filamentous species
S7	Leptolyngbya boryana An.	Safi / Reservoir / 08‑2004	Attached to the substrate of the banks reservoir
S8	Leptolyngbya lurida (Sku.) An.	Ourika / Stream / 11‑2006	Attached to substrate of the edge of the stream
S9	Leptolyngbya sp.	Sebkha Zima / Lac / 07‑2005	Attached to other filamentous species
S10	Limnothrix rosea (Ute.) Mef.	Oued Tassaout / River / 06‑2004	Attached to gravels in the calm zones of river
S11	Lyngbya hieronymusii Lemm.	Imfout / Reservoir / 07‑2004	Attached to the substrate of the banks or with other filamentous species
S12	Microcystis wesembergii. (Kom.) Kom.	Tigelmamines / Lac /10‑2005	Float on the surface of the water mass forming scum or bloom
S13	Microcystis aeruginosa(Kütz.) Kütz.	L. Takerkoust / Reservoir/ 07‑2003	Float on the surface of the water mass forming scum or bloom
S14	Oscillatoria limosa Ag.	My. Youssef / Reservoir / 10‑2004	In the water mass and other filamentous species
S15	Phormidium articulatum Clau.	Bouznika / River / 10‑2004	Attached to muddy substrate or with other filamentous species
S16	Phormidium chalybeum (Mer.ex Gom.) An. et Kom.	Bouznika / River / 10‑2004	Attached to muddy substrate or with other filamentous species
S17	Phormidium Koprophilum (Sku.) An.	Oued Mellah / River / 07‑2004	Attached to muddy substrate or with other filamentous species
S18	Phormidium lusitanicum Sam.	Oued Cherat / Benslimane / 07‑2004	Attached to muddy substrate or with other filamentous species
S19	Phormidium paulsenianum (Boy.) Nov.	Oued Mellah / River / 07‑2004	Attached to substrate of the edge of the river with other filamentous species
S20	Phormidium sp.	ElKansera / Reservoir / 10‑2005	Attached to other filamentous species forming floating mats
S21	Phormidium subfuscum Kütz.	Sidi Daouad / Canal, Marrakech / 10‑2004	Attached with other filamentous on the edge of canal
S22	Planktothrix agardhii(Gom.) An. et Kom.	M. Eddahbi / Reservoir/ 10‑2004	Float on the surface of the water mass forming scum or bloom
S23	Pseudanabaena biceps Bour.	Ourika / Stream / 11‑2005	Attached to substrate of the edge of the stream
S24	Pseudanabaena catenata Laut.	L. Takerkoust / Reservoir / 07‑2003	In the water mass and other filamentous benthic species
S25	Pseudanabaena galeata Böch.	Lac Zima / Lac / 07‑2003	Attached to other filamentous species forming floating mats
S26	Pseudanabaena lonchoides An.	M. Eddahbi / Reservoir / 10‑2004	In the water mass, non forming bloom
S27	Pseudanabaena papillaterminata Kuk.	M. Eddahbi / Reservoir / 10‑2004	In the water mass, non forming bloom
S28	Synechocystis minuscula Vorn.	Azigza / Lac / 10‑2005	In the water mass, non forming bloom
S29	Synechocystis salina Wis.	Bou Regreg / estuary / 07‑2004	In the water mass, non forming bloom

Among all isolated strains, two showed a positive toxicity (Pseud. C. S24 and Pseud. G. S25) with respective LD⁵⁰ of 292.8 mg•kg^‑1 DW and 1,192.7 mg•kg^‑1 DW. After i.p mice bioassay, the observed poisoning signs are similar to those observed with Microcystis, MCs producing extract, except for the strain (Pseud. G. S25), which present severe diarrhea and a longer survival time (5 to 7 h), more than one to two hours specific for hepatotoxins. Although the positive mice biotest, the HPLC-PDA analysis showed a negative result for detecting MCs in both isolated strains (Pseud. C. S24 and Pseud. G. S25), whereas, for the three others (Mic. TAK. S13, Plank. A. S22 and Pseud. B. S23) are confirmed as MCs producers with a detection of five MCs variants (see Table 3 for details). According to HPLC-PDA results, only three strains, Mic. TAK. S13, Plank. A. S22 and Pseud. B. S23, among the 29 tested species, are confirmed as MCs producers.

4. Discussion

The widening of the survey in various Moroccan aquatic environments allows a thorough knowledge of the diversity of cyanobacteria. The priority to evaluate the potential risk of these micro-organisms led us to couple survey with toxicological assessment of blooms and/or mats and isolated strains.

4.1 Cyanobacterial biodiversity

The taxonomic richness of cyanobacteria was remarkably increased (Figure 1). This is due to several factors like the diversity of the Moroccan aquatic environments and the widening of the study scale, in the context of the research program on toxic cyanobacteria initiated in the last years.

The predominance of taxa of Oscillatoriales (Figure 2) in these aquatic environments is not a surprise because several works all over the world already indicated this predominance, often in mesotrophic or eutrophic waterbodies (BRANCO et al., 2003 ; MUR et al., 1999 ; SCHEFFER et al., 1997). Microcystis genus is the most common taxon in all lakes and reservoirs of Morocco (LOUDIKI et al., 2002). Other species of cyanobacteria are cosmopolitan and are frequent in fresh waters and mineralized water.

In lakes and reservoirs (Figure 4), it seems that the stability of water column and nutrient increasing input are in favour of higher biodiversity and cyanobacteria massive proliferation. In contrast, in streams and mountain springs, the diversity of the cyanobacteria is relatively low (12%) with the predominance of some particular genera forming often mat (Nostoc, Lyngbya).

4.2 Toxicological assessment

4.2.1 Cyanobacteria bloom or mat‑forming

The toxicity and cyanobacterial toxins (MCs) content of natural blooms and mats have been analyzed in several Moroccan freshwaters. The positive toxicity and detection of MCs were obtained in three reservoirs: Lalla Takerkoust, Mansour Eddahbi and Almassira, and in the natural lake Dayet Erroumi. In all these localities, although the M. aeruginosa is the predominant species responsible for blooms, the toxicity determined both by mice and Artemia bioassays, is different (Table 3). With respect to bloom toxicity level and according to classifications suggested by Lawtonet al., (1994) ; Mic. A. TAK bloom exhibiting an LD⁵⁰ = 176.8 mg•kg^‑1 is highly toxic and both for Mic. A. MED (LD⁵⁰ = 352 mg•kg^‑1) and Mic. A. ALM blooms (LD⁵⁰ = 829 mg•kg^‑1) are moderately toxic.

The toxicity of the Mic. A. TAK and Mic. A. MED bloom could be regarded as higher than that of a bloom from the brackish Moroccan Oued Mellah lake (LD⁵⁰ = 502 mg•kg^‑1), whose dominant species was Microcystis ichtyoblabe (SABOUR et al., 2002).

According to the mouse biotest, the Mic. A. ALM bloom was only moderately toxic (829 mg•kg^‑1). The toxicity is about six times lower than that observed (142  mg•kg^‑1) in a Microcystis bloom from the same reservoir in 1999 (OUDRA et al., 2001b). The toxicity dissimilarity could be attributed to the difference in the growth phase of the bloom and also to the environmental conditions during blooms occurrence. To this respect, SABOUR et al. (2002), while studying Microcystis bloom in Oued Mellah reservoir, observed a change in the mouse LD⁵⁰ from 518 to 1924 mg•kg^‑1, corresponding respectively to the exponential and decline growth phase of the bloom. During the decline growth phase, most of the MCs should be detected in the water, since these toxins are released to the medium after cell disruption. The highest toxicity reported corresponds to a M. aeruginosa bloom from the Lalla Takerkoust reservoir (OUDRA et al., 2002a).

A similar highly variable toxicity in Microcystis blooms from a Mediterranean reservoir has been reported for Kastoria lake, in Greece, with a LD⁵⁰ ranging from 40 to 1,500 mg•kg^‑1 (COOK et al., 2004). In other cases, as in Portugal, in a previous work carried out during 1989-1992, the studied blooms, dominated by Microcystis, had a LD⁵⁰ remaining lower than 700 mg•kg^‑1 (VASCONSELOS, 2001).

According to the Artemia biotest, the Mic. A. TAK, Mic. A. MED and Mic. A. ALM blooms are classified as toxic. But a significant difference between them is observed, 24‑h LC⁵⁰ = 1.41 mg•mL^‑1 for Mic. A. TAK bloom, 24‑h LC⁵⁰ = 1.71 mg•mL^‑1 for Mic. A. MED bloom and 4.34 mg•mL^‑1 for Mic. A. ALM bloom.

As with the mouse bioassay, these toxicity values are higher than those previously reported for the Microcystis bloom in Oued  Mellah reservoir, 24‑h LC⁵⁰ of 6‑46 mg•mL^‑1 (SABOUR et al., 2002).

In general, our results with Artemia bioassay clearly confirm that this bioassay can be used as an alternative test to evaluate cyanobacteria toxicity.

MCs concentration in Mic. A. TAK, Mic. A. MED, Mic. A. ALM and Mic. A. blooms were 64.4, 13.94, 9.9 and 1.87 µg•g^‑1 eq MC-LR respectively.

The MCs concentration in all blooms is higher than that observed before in the ALM (OUDRA et al., 2001b) and Oued Mellah (SABOUR et al., 2002) reservoirs. However, the content is very low compared with that reported for Microcystis blooms of diverse origins such as the Moroccan Lalla Takerkoust lake (8.8 mg•g^‑1, OUDRA et al., 2001a), various Portuguese reservoirs (1-7.1 mg•g^‑1; VASCONCELOS, 2001), the Spanish Santillana reservoir (13.5 mg•g^‑1; PADILLA et al., 2006) and some French reservoirs (0.07-3.97 mg•g^‑1, VEZIE et al., 1997).

The toxicity evaluation by mouse biotest of mat-forming N. Muscorum (Nos. M) and L. attenuata (Lyn. A1,2) in Ourika and Oukaïmeden streams was positive (Table 3). The symptoms were also similar to those of hepato-toxicosis. In one of these mats, MCs were detected (Table 2). This could be explained by the presence of small quantities undetectable by HPLC or the presence of other types of toxins. Indeed, for a sample of Nostocmuscurum bloom collected in April, 1999 from OK stream, the MCs concentration was about 229.4 µg, equivalent MC-LR•g^‑1 DW and an LD⁵⁰ of 125 mg•kg^‑1 (OUDRA et al., 2008). Over the world, a few studies related to benthic toxic cyanobacteria that could be cited (HITZFELD et al., 2000 ; IZAGUIRRE et al., 2007 ; MEZ et al., 1998 ; OUDRA et al., 2008 et ZAKARIA et al., 2006). According to Zakariaetal. (2006), these benthic cyanobacteria could produce anatoxins, saxitoxins or microcystins.

With respect to MCs characterization, the cyanobacterial natural bloom shows a net difference in composition: the ALM bloom producing only MC-LR is the less diversified one, whereas the TAK bloom presents four MCs variants and the most diversified blooms are MED and DE Microcystis blooms with five MCs variants (Table 3).

The predominance of hydrophobic variant MC‑WR (known as slightly toxic) in Mic. A. TAK bloom and Mic. A. MED bloom constitutes an exceptional case, not only for the studied reservoirs but also for Morocco and, to our knowledge, for the Mediterranean region.

The predominance of MC-RR over the other variants has been reported for M. aeruginosa in the Philippines (CUVIN-ARALAR, 2002). Predominance of MC‑LR, MC‑RR and MC‑YR was also reported for some countries of the Mediterranean region, like Egypt (ABDEL-RAHMAN et al., 1993), Morocco (OUDRA et al., 2001a), Portugal (VASCONSELOS, 2001), Spain ( PADILLA et al., 2006) and Algeria (NASRI et al., 2004).

The presence of only one variant of MCs in Mic. A. ALM bloom suggests that several toxic cyanobacteria often contain one major toxin. They may also contain several other toxins but in minor quantities (HARADA et al., 1991 ; SHIRAI et al., 1991). In Portugal, microcystin LR is usually the major toxin with a dominance ranging from 45.5 to 99.8% (VASCONCELOS, 2001).

The predominance of a variant may be related to the difference in strain composition dominating the bloom because microcystin-producing and non-producing strains can coexist in populations of cyanobacteria (ROHRLACK, 2001 ; KURMAYER et al., 2002 ; VEZIE et al., 1998).

4.2.2 Cyanobacteria isolate strains

The results obtained show that Pseud. G. S25 strain has a relatively high toxicity (292.8 mg•kg^‑1 DW) while Pseud. C. S24 strain reveals a medium toxicity (1,192.74 mg•kg^‑1).

In spite of their hepatotoxicity, MCs were not detected by HPLC in Pseud. C. S24 or Pseud. G. S25 strains. This can be explained by the presence of small undetectable quantities of MCs or other types of cyanotoxins. In fact, a species morphologically similar to Pseud. G. S25, determined as Pseudanabaena galeata, which was already isolated from the experimental wastewater stabilisation ponds of Marrakech, was confirmed as MCs producer (OUDRA et al., 2002b).

In addition, the mat‑forming Pseud. B. S23 strain, isolated from a High‑Atlas stream, was identified for the first time as a MCs producer strain in Morocco. The complete HPLC‑MS characterization of the unknown microcystin variants is in progress. Related to producing toxins amount, MCs content in the Mic. TAK. S13 strain (9.44 µg•g^‑1 eq MC‑LR) was about three times higher than in the Plank. A. S22 strain (3.19 µg•g^‑1 eq MC‑LR) and about 34 times than in Pseud. B. S23 (0.28 µg•g^‑1 eq MC-LR) (Table 3).

The comparison of the diversity of MCs in Mic. A. TAK bloom and their isolated strains, shows the appearance of MC‑YR variant under normal conditions of culture at the laboratory. The variation of the percentage (quotas) and type of MCs is related to the conditions of the culture medium. MC-YR was already observed in Mic. A. TAK bloom (OUDRA et al., 2002a). These results indicate that blooms have the potential to contain these toxins or others. However, the Plank. A. S22 strain produced only one variant (MC‑RR). This strain is known for the production of MC‑RR and demethylmicrocystins variants (FASTNER et al., 1998 ; MESSINEO et al., 2006 ; SIVONEN et JONES 1999).

Though no sign of toxicity or presence of toxins was revealed, the strains S1-S12, S14-S21, S26-S29 remain potentially toxic, knowing that some of them are already known as toxic such as S1‑S12. This observation is supported by the fact that we tested only the cultures when it is often observed that blooms are more toxic. Culture conditions (temperature, nutrients, trace metals, etc.) can act on the strains nature and on their toxinology (KOTAK et al., 2000).

Only a genetic approach of some natural strain isolates can reveal this potential in order to draw up, firstly, a red list of the high-risk zones likely to have a proliferation of the toxic cyanobacteria, and secondly, to exploit biotechnologically the non‑toxic strains.

5. Conclusion

The taxonomic study provided a recent, updated and relatively complete national inventory of the cyanobacteria, whose current number is now 578 taxa described for the first time in Morocco. Genetic analysis is necessary to determine their taxonomic identity. These results constitute a primary database for the establishment of a national cyanobacteria culture collection in Morocco.

The toxicological results show the presence and the toxicity of the MCs in four natural Microcystis blooms and three isolates. Five MCs variants are detected.

Further research, focused on the presence of other variants of cyanobacterial toxins, seems necessary. In this direction, a future approach will focus on the detection of the toxigenic genes.

This study has the merit to bring additions to previous knowledge and to map toxic strains that can generate harmful effects on human and environmental health. Thus, it could be very useful for any Moroccan water quality monitoring and sanitary risk survey program.