There is now ample evidence of the impact of the recent climate change and anthropogenic activities on different saline lake ecosystems. All over the world salt lakes are threatened by climate change, water diversions upstream for agricultural purposes, watershed changes, introduction of alien species, etc. that resulted in catastrophic changes in the different lakes around the world(O’Reilly et al., 2003; Smol et al., 2005; Anneville et al., 2007; Abbaspour et al., 2012; Shadrin and Anufriieva, 2013b). Different negative consequences for human communities resulted from this. In our world some changes are a natural result of the global climate system variability, but the responsibility for other changes is entirely due to human activity. We need to know how to separate these two groups of the causes of the salt lake changes to predict and mitigate the negative results. To do so we need at first to have long-term monitoring data and perform a comparative analysis for the different lakes.

The Fayum(Faiyum), a natural depression(approximately 80 km southwest of Cairo), extending over 12 000 km ², was formed by wind erosion ca 1.8 million years ago(Ball, 1939). The depression is bounded by s and y hills, broken in the south, where Canal Bahr Yousef enters the depression. In the northern Fayum Lake Qarun, the third largest lake in Egypt is located(Fig. 1). Archaeologists have shown that the modern Lake Qarun is a shrunken remnant of Lake Moeris, known from since 450 B.C. Herodotus, who visited Egypt, saw a large water expanse, Lake Moeris, assumed artificial(Ball, 1939). At that time the lake was 75 m deep(20 m above sea level), covered more than 2 000 km ², and was maintained by a seasonal supply of Nile water(Ball, 1939; Shafei, 1960). Different studies showed that the lake level has varied from relatively high levels in its early history to lower levels in later years, although there have been numerous cycles in the water level over the past 7 000 years(Nicoll, 2004; Baioumy et al., 2010). Lake Qarun has been profoundly affected by a combination of human activities and climatic changes during the past 5 000 years(Shafei, 1960; Nicoll, 2004; Baioumy et al., 2010). During the last 2 000 years there were several periods of low water level; from 17 m below sea level or lower in the third century A.D to 30 m below sea level in 1245 A.D. when Nabulsigave a description of the Faiyum area; in 1805–1848, when Mohamed Aliruled Egypt, it averaged 40 m below sea level; in 1933–1934, 45– 46 m below sea level(Baioumy et al., 2010). The drop in the lake level and the accumulation of salts led to salinity increase; Lake Qarun was only slightly brackish until about 1884. Salinity increased from 8.5 g/L in 1905 to 38.0 g/L in 1980 due to two factors: natural climate variability and human activity(El Shabrawy and Dumont, 2009; Baioumy et al., 2010). To reduce the salinity increase in the lake, the Egyptian Company of Salts and Minerals(EMISAL)was established on the southern coast of Lake Qarun in 1986 to extract salts and minerals(EMISAL, 1996). Its activity now contributes to salt balance regulation in the lake. The current salt balance in the lake(annual average)includes: 1. salt discharge via the main drains of 419.56 million kg; 2. groundwater provides 70.36 million kg; 3. salts extracted by the EMISAL plant amount to 416 million kg. Concerning the salt budget, during recent years(up to 2007)Lake Qarun accumulated 70–85 million kg per year. This may lead to a salinity increase of 0.07 g/L per year(Abd Ellah, 2009). The question is whether such a salinity increase may lead to catastrophic changes in the lake’s plankton.

Fig. 1 Sampling stations in Lake Qarun during August 2011 survey

Figure options

Zooplankton plays an important ecological role in aquatic ecosystems. It occupies different ecological niches in aquatic food webs, contributes to biologicial element cycles, and transfers organic matter from primary producers to higher secondary consumers like fish. The distribution of zooplankton responds to eutrophication, pollution, and global warming(Gannon and Stemberger, 1978; Meshram, 2005). Among zooplankton, rotifer populations with their high turnover rates are particularly sensitive to water quality changes. Different studies have shown that eutrophication affects zooplankton composition, shifting the dominance from large species(Copepoda)to smaller ones(Rotifera)(Gannon and Stemberger, 1978; Premazzi and Chiaudani, 1992).

Wimpenny and Titterington(1936)reported that the zooplankton of Lake Qarun consisted mainly of brackish water species, at that time dominated by Arctodiaptomus salinus(Daday, 1885)(Copepoda) and Moina salina Daday, 1888(Cladocera). Naguib(1958) and Girgis(1959), twenty years after, noticed that there was no more evidence of A . salinus and M . salina, while the marine neritic copepod Paracartia latisetosa(Kritchagin, 1873)was detected in high numbers. P . latisetosa was first recorded in 1930(Naguib, 1961). To maintain fisheries in the lake, mullet, eels and sole have been periodically introduced from the Mediterranean since 1928(Naguib, 1961). Different plankton and benthic animals were occasionally transported to the lake during these fish introductions. Girgis(1959)wrote that the permanent zooplankton in Lake Qarun was composed mainly of copepods(P . latisetosa) and Protozoa. Only marine zooplankters, which were transported with an introduction of marine fishes, inhabit the lake now(Abdel-Malek and Ishak, 1980; Dowidar, 1981; El- Shabrawy, 2001; El Shabrawy and Belmonte, 2004; Mageed, 2005; Khalifa and El-Shabrawy, 2007).

During last 100 years the huge changes in structure and functioning of biota in the lake occurred. At first glance we may explain these changes by the salinity increase; many researchers did so(Naguib, 1958; Abdel-Malek and Ishak, 1980; Ishak and Abdel- Malek, 1980; Mageed, 2005; El Shabrawy and Dumont, 2009). It must be asked whether this explanation is correct and also applies to zooplankton. We still do not have enough knowledge to underst and the interplay of various factors determining the structure and dynamics of zooplankton in the lake. The aims of our work are to provide new data on the lake zooplankton, to discuss the long-term changes of zooplankton, and to evaluate whether salinity change may be a main driver of these changes.

Lake Qarun is a closed saline lake in the northern part of the El-Fayum Depression(Middle Egypt, at the margin of the Nile Valley). The studied lake is located at 29°30'N, 30°30'E(Fig. 1) and is 43–43.5 m below sea level(Abdel-Satar et al., 2010). The lake length from east to west is about 40 km, and the maximal width is about 6.7 km. The lake has a surface area of 243 km ² and a volume of 924 million m ³(El Shabrawy and Dumont, 2009). The maximum depth(~8.3 m)is in the northwest part. The non-irrigated northern shores of the lake are actually devoid of vegetation and mark the beginning of the Western Egyptian Desert. The lake, located 320 km south of the Mediterranean coast of Egypt, has no connection to the sea, and is sustained directly by the Nile River via the Bahr Yussef canal. This is main source of water for the lake. The lake receives fresh water from the Nile since early Pharaonic times, i.e., before 2500 B.C.(Hassan, 1986). Terraced lands allow a use of masonry weirs for the distribution of irrigation waters. Lake Qarun collects agricultural drainage water through two drains—the El-Bats and El-Wadi systems. In 1952–1954 the annual average amount of drainage waters entering the lake were about 349.2 million m ³(Naguib, 1958) and in 2 000, about 400 million m ³(Abdel-Satar et al., 2010).

Zooplankton samples were collected by vertical hauls from bottom to surface at 10 stations in August 2011(Fig. 1); from the previous studies(El-Shabrawy, 2001; Khalifa and El-Shabrawy, 2007)we recognized that they represent the different microhabitats of plankton in the lake. Vertical plankton net with mesh size of 55 μm was used. In addition, qualitative samples were taken for detection of rare and sporadic species. The samples were immediately preserved in 4% neutral formalin solution. Identifications of zooplankton were made using an Olympus BX50 compound microscope.

Water temperature(°C), pH and electrical conductivity(EC, μS/cm)were in-situ measured using Hydrolab model MultiSet 430iWTW. Transparency was measured using a white/black SecchiDisk(20 cm in diameter). Total solids(TS)were measured by evaporating a known volume of well mixed sample. Total dissolved solids/salinity(TDS)was determined by filtering a volume of sample through a glass fiber filter(GF/C), and a known volume of filtrate was evaporated at 105°C. Total suspended solids(TSS)equal the difference between TS and TDS. Dissolved oxygen(DO, mg/L)was assayed using a modified Winkler method. Biological oxygen dem and (BOD)was determined by using the 5 days incubation method. Chemical oxygen dem and (COD)was carried out using the potassium permanganate method. Water alkalinity was determined immediately after sampling collection using phenolphthalein and methyl orange indicators. Sulfate was determined using a turbidimetric method. Calcium and magnesium were detected by using a complexemetry method by direct titration using EDTA solution. Ammonia was determined by the phenate method. Nitrite was determined using a colorimetric method with formation of a reddish purple azo-dye. Nitrate was measured as nitrite after cadmium reduction. Orthophosphate was estimated by using the ascorbic acid-molybdate method. Reactive silicate was determined using the molybdate method. Total nitrogen(TN) and total phosphorus(TP)were measured as nitrate and orthophosphate respectively, after persulfate digestion(APHA, 1995).

Data were subjected to st and ard statistical processing(Sokal and Rohlf, 1995). Variability of parameters in the samples was qualified by the coefficient of variability(CV). Pair coefficients of correlations(R)were calculated in Excel for all possible pairs. Significance of differences in average values was evaluated by Student’s t-test, and the confidence level of correlation coefficients(P)was determined by comparison with parameter critical values(P ≤0.05)(Müller et al., 1979). The STATISTICA software package(Version 6.0; Statsoft, Inc.)was utilized to calculate Euclidean distances between stations and tree clustering using all abiotic and zooplankton measured parameters.

The values of studied physico-chemical parameters for all stations are given in Table 1. There is no thermal stratification in the lake because it is shallow. BOD and COD did not correlate. COD significantly positively correlated with salinity(R =0.667; P =0.05). The proportion between BOD and COD(BOD/COD)slightly fl uctuated; it ranged from 0.27(station 10)to 0.37(station 1). This proportion demonstrated significant positive correlation with NH ₄(R =0.807; P =0.01). Salinity had a significant negative correlation with nutrient concentrations and transparency in the lake: with NH ₄(R =-0.823; P =0.01), NO ₃(R =-0.926; P =0.005), PO ₄(R =-0.640; P =0.05), SiO ₄(R =-0.75; P =0.05), transparency(R =-0.862; P =0.01). Other calculated correlation coefficients were insignificant. Comparison of the CVs for different abiotic parameters showed different levels of spatial variability. Cluster analysis with a use of all measured abiotic parameters(Fig. 2)showed that stations 1, 2 and 7 compose the first group and all other stations the second group.

Table 1 Spatial variability of physico-chemical parameters in Lake Qarun (August 2011)

Table options

Fig. 2 Dendrogram of the stations based on abiotic parameters in Lake Qarun, using group-average linking of the Euclidean distance coeffi cient

Figure options

Fifteen holozooplankton species, with total average abundance of 595 000 ind./m ³(CV=0.274), were identified during the present study(Table 2). The highest abundance was observed at station 5(845 000 ind./m ³), the lowest at station 1(294 000 ind./ m ³). CV value shows that distribution of total zooplankton abundance in the lake was closer to the uniform distribution. CV for Protozoa was 0.65, 0.84 for Rotifera, 0.93 for Copepoda, and 0.63 for Meroplankton. Such CV values show that distribution of single zooplankton groups in the lake was closer to the uneven distribution than in total zooplankton.

Rotifera was a dominant group, particularly at all western side stations, comprising 49.7% of total zooplankton abundance(Table 2). Station 10(the westernmost station)maintained the highest density of 606 000 ind./m ³, while the lowest value of 33 000 ind./m ³ was recorded at station 4. Coefficients of correlation between abundance of different rotifer species and salinity were low and insignificant. B . cf. rotundiformis, S . cf. kitina and Bdelloidea(unidentified)were the most common species. B . cf. rotundiformis had the same trends of its distribution in the lake as total Rotifera, while the density of S . cf. kitina reached a maximum(41 000 ind./m ³)at station 5(Table 3). The logarithm of B . cf. rotundiformis density had a significant positive correlation with temperature(R =0.834; P =0.01). This relationship may be approximated by the equation:

Click browse formula

Table 2 Average species abundance (10 ³ ind./m ³ ) at different stations in Lake Qarun (August 2011)

Table options

Table 3 Long-term changes of physico-chemical parameters in Lake Qarun

Table options

The proportion between the densities of B . cf. rotundiformis(B) and S . cf. kitina(S), in which B/S was infl uenced by temperature(R =0.906; P =0.001); this may be approximated by:

Click browse formula

Copepoda occupied the second dominant position—45.4% of total zooplankton density. Two species dominated; P . latisetosa(Calanoida)was a dominant species in the zooplankton at all stations, Canuella sp.(Harpacticoida)was subdominant in 5 stations(3, 4, 5, 7, and 10). A . panamensis was collected only at station 4. Nauplius stages were abundant in the eastern and middle parts of Lake Qarun during this season, while copepodid and adult stages of P . latisetosa were most abundant at the westernmost station(station 10)(Table 2). There was a negative significant correlation between the logarithms of total densities of Copepoda and Rotifera(R =0.869; P =0.001); this relation may be approximated by the equation:

Click browse formula

Coefficients of correlation showed that there was a significant negative infl uence of temperature(R = -0.845; P =0.005) and PO ₄(R =-0.574; P =0.05)on the proportion between Copepoda(C) and Rotifera(R)densities(R/C). Meroplankton contributed 3.1% of the total zooplankton abundance. It was represented by six groups and reached a maximum at station 5, while stations 6 and 9 sustained the lowest abundance(Table 2).

Cluster analysis, with a use of data on total zooplankton number and abundance of all single species, demonstrated that the stations grouped as follows: the first group included stations 1, 2, and 3; the second group: stations 4 and 5; the third: all other stations(Fig. 3). This composition is very different from that given by clustering based on abiotic parameters. When we performed cluster analysis separately with data on protists, rotifers, copepods or meroplankton only, we obtained four different dendrograms. A tree of clusters for copepods is given in Fig. 4.

Fig. 3 Dendrogram of the stations based on total zooplankton parameters in Lake Qarun, using group-average linking of the Euclidean distance coeffi cient

Figure options

Fig. 4 Dendrogram of the stations based on Copepoda parameters in Lake Qarun, using group-average linking of the Euclidean distance coeffi cient

Figure options

Introduction of aliens continues. The last alien species introduced was the warty comb jelly Mnemiopsis leidyi A. Agassiz, 1865. It was first reported at Lake Qarun in March 2014. Large numbers of the ctenophore were reported by many fishermen and by the NIOF research team. Probably, M . leidyi was accidentally introduced in the lake through mullet fry transportation. M . leidyi was recorded also in a second lake of the Fayum depression(Wadiel Rayan), where large swarms had been recorded(fishermen’s observations).

Stations(1 and 7)with the lowest salinity and the highest nutrient concentrations and turbidity were closest to the discharge of waters from the El-Bats and El-Wadidrainage systems. This has been noted before(Mageed, 2005). Girgis(1980)suggested that the turbidity in Lake Qarun was mainly caused by discharge of agricultural drainage waters. In our study BOD was lower than COD values; this agrees with Goldman(1972), who reported that the BOD of wastewater was lower than COD. Organic substances entering Lake Qarun from its watershed are mainly produced by the decomposition of domestic wastes and by and particulate organic matter discharged through two drains and several small drains present around the lake(Siliem, 1993).

Great differences in variability of the studied parameters enable us to conclude that different external and internal factors infl uence the variability of these parameters. Climate changes at first drive the changes of the factors, which have low spatial variability, but local factors, including an anthropogenic impact, are more important to determine the values of the factors with high spatial variability.

Available data on salinity since 1901(Ishak and Abdel-Malek, 1980; Abd Ellah, 2009; El Shabrawy and Dumnont, 2009; Baioumy et al., 2010)showed that there was a strong salinity increase since 1901(12 g/L)to approximately 34–39 g/L in 1995–2000. Later there were salinity fl uctuations and only a slightly positive trend(Table 3). Long-term studies on the lake(Naguib, 1958; Soliman, 1990; Ali, 2002; Sabae and Ali, 2004; Abdel-Satar et al., 2010)provide evidence for a current increase of concentrations of the major nutrients due to a growing impact of water discharge from drains. The mean concentration of the major nutrients(nitrate, nitrite and orthophosphate)gradually increased from ³5, 0.16 and 0.38 μ g/L, respectively in 1953–1955 to 113, 16.4, and 30.26 μ g/L in 2011(Table 4). Organic carbon in the lake water(as COD)fluctuated during last decades: 11.99 mg/L in 1989; 28.1 mg/L in 1999–2000; 11.99 mg/L in 2006 and 15.05 mg/L in 2011(Table 4), we cannot identify the causes of this fluctuation.

In August 2011 the species composition of zooplankton in the lake was close to that previously reported(Abdel-Malek and Ishak, 1980; El-Shabrawy, 2001; Mageed, 2005; El Shabrawy and Dumont, 2009). A comparison with those data showed that some decrease of species diversity occurred. Some species disappeared from the lake; some protists(G . infl ata, C . glaucoma, Arcella sp.)were noted in the lake in earlier times. The distribution of rotifer density in the lake found by us agrees with data of El- Shabrawy(2001) and Mageed(2005), in contrast with data from Abdel-Malek and Ishak(1980)who mentioned that rotifers were significantly higher at the eastern than the western area of the lake. Wind direction drives a spatial distribution of zooplankton in shallow lakes. Differences in a spatial distribution of zooplankton in different seasons and years may be caused by changes of prevailing wind direction. Some authors(Abdel-Malek and Ishak, 1980; El-Shabrawy, 2001; Mageed, 2005)supposed that salinity is a limiting factor in the distribution of rotifers, but we did not find such correlation in 2011. El-Shabrawy(2001)previously recorded a positive correlation between B . cf. rotundiformis and water temperature, salinity and nitrate in the lake. No significant correlations between B . cf. rotundiformis and salinity, nitrate and phosphate were found in this study. P . latisetosa was also previously dominant in Lake Qarun(Abdel-Malek and Ishak, 1980; Dowidar and El-Nady, 1982; El Shabrawy and Belmonte, 2004).

Clusters of the stations in Lake Qarun made separately for total zooplankton, protozoa, rotifers and copepods were different from each other. Similar results were obtained also in earlier periods: for data collected in 2003 clustering of the stations in Lake Qarun separately according to protozoans, rotifers and copepod was performed by Mageed(2005); similar differences in the dendrograms were observed. This means that the composition of abiotic parameters does not directly determine spatial differences in zooplankton structure, and that there are no unambiguous links between the different groups of organisms in the zooplankton. Irregularities such as strong winds or other occasional events also take part in ruling of plankton structure.

During last decades the total zooplankton density and its composition have varied in a wide range; the average total density was 30 000 ind./m ³ in 1974– 1977; 356 125 ind./m ³ in 1989; 534 000 ind./m ³ in 1994–1995 and from 965 000 to 1 452 000 ind./m ³ in 2006(Ahmed, 1994; El Shabrawy and Belmonte, 2004; Mageed, 2005; Khalifa and El-Shabrawy, 2007). The average total zooplankton abundance for the 1974–2011 period was 503 000 ind./m ³(st and ard deviation=355 000, CV=0.71)fl uctuating from ³0 000 ind./m ³ to more than 1 000 000 ind./m ³ . We did not find a significant correlation between changes of zooplankton abundance and salinity, there was only insignificant positive trend. The average total zooplankton abundance in August 2011(595 000 /m ³)was very close to average in 1974–2011. The contribution of different zooplankton groups to the total abundance also was different in different years and seasons(Ahmed, 1994; El Shabrawy and Belmonte, 2004; Mageed, 2005; Khalifa and El- Shabrawy, 2007). There is no single trend of these changes.

A comparison of the spatial changes in August 2011(Tables 1, 2)with seasonal and inter-annual variations(Table 3 and above in text)showed that spatial variability of all studied parameters, including zooplankton ones, was essentially less than seasonal and inter-annual temporal variability of these parameters. However, summer salinity variability during the last decades(from ³0 g/L to 39 g/L)was very close to a range of spatial variability in summer 2011(from 21 g/L to 38 g/L). We may conclude that salinity fl uctuations since at least 1955 did not have a significant direct impact on the composition and abundance of zooplankton in the lake because no significant correlation was observed in both cases. We assume that the disappearance of A . salinus and M . salina in the plankton of Lake Qarun, which was explained previously by salinity increase(Naguib, 1958), could not be due to observed salinity increase; halotolerance of those species is very high(Shadrin and Anufriieva, 2013a, b). In North Africa development of M . salina was observed in hypersaline waters at salinities up to 225 g/L(Amarouayache et al., 2012). Similar observations were reported from other parts of the world(García and Niell, 1993; Shadrin and Anufriieva, 2013b). A . salinus also is known from hypersaline waters in North Africa(Rokneddine and Chentoufi, 2004), as well as on other continents(Shadrin and Anufriieva, 2013a). A salinity increase in water bodies of arid and semi-arid zones may be expected to continue because there is a trend of global warming. We may suppose that a further increase of salinity to 50–60 g/L may lead to a reappearance of M . salina and A . salinus in the plankton of Lake Qarun. Resting stages of these two species may survive for a long time in the lake sediments and be transported to the lake by birds(Sánchez et al., 2007). However, taking into account the current trend of the salinity increase(see above), such salinity would be observed only in the 23rd century.

Analyzing all data we suppose that from the middle of the 19th century until 1928 there was a first stage of a modern biotic transformation in Lake Qarun when a salinity increase was the main cause of the changes. Now there is a marine community in the lake, which continues to change also today. One of main drivers for this is a regular introduction and a pressure of alien species on the existent species. Eutrophication also plays an important role in the changes of zooplankton in the lake. As a result of eutrophication, a recurrent microalgae blooming phenomenon was recorded in the last years(2008 to 2012)in the autumn or winter seasons(Abou El-Geit et al., 2013). Some observed changes in plankton may be explained by episodic microalgae blooms and mass mortality. Discussing the reason of changes in Lake Qarun, we cannot forget its pollution by heavy metals, pesticides, and other pollutants(Mansour and Sidky, 2003). As Lake Qarun is a terminal lake with an intensive evaporation; salts, heavy metals and pesticides, carried by agricultural and municipal drainage waters, accumulate in the lake, worsening the situation. We conclude that many factors contribute to the current changes in Lake Qarun plankton with different chain, domino, top-down and up-down effects. Therefore it is very difficult task to quantitatively separate the causes of the current long-term changes in the ecosystem. The introduction of M . leidyi into the lake may lead to a start of a new stage of the biotic changes in Lake Qarun, when eutrophication and a population dynamics of this ctenophore will be the main drivers of the ecosystem change. This ctenophore was introduced in the Black Sea in the 1980s, which had catastrophic results for the Black Sea ecosystem and its fishery industry(Gomoiu et al., 2002). It was supposed that eutrophication and ecosystem destabilization promoted the successful occupation of the Black Sea and other seas by M . leidyi.

All the above information supports a general conclusion that there is no single abiotic parameter that determines the composition and density of zooplankton in Lake Qarun. The composition of abiotic factors also is not a strong determinant of zooplankton structure in the lake. The increase in salinity/nutrient concentration may have an impact on biotic interactions in the plankton, and this may result in changes of species composition. We agree with Williams(1998)that salinity as well as other abiotic factors does not act a major determinant of species occurrence: the major determinants probably included stochasticity, predation, food availability, competition, other forms of biological interaction, and interactions between particular physical and chemical factors. Remembering the multiplicity of the alternative states of the salt lake ecosystems(Shadrin, 2013), we cannot make a strong unequivocal forecast of future changes in Lake Qarun. The chance to make such a forecast is further decreasing due to increased climate instability(McElroy and Baker, 2012). Violating natural l and scape and ecological connectivity we multiply the problems, including also a predictivity of future changes. Study of spatial heterogeneity in the lakes may promote a better underst and ing of the interconnection of different ecosystem parameters and temporal variability of the lake ecosystems.