Zooplankton community structure in relation to environmental variables in a tropical endorheic Soda Lake Shala, Ethiopia

Abstract

Lake Shala, endorheic Soda Lake in the Ethiopian Riftvally Lake, supports an enormous biodiversity, with a large human population and economic development. However soda lakes are under several ecological changes, with unexpected shifts in biodiversity composition and physicochemical changes. This study focuses on assessing the community structure of zooplankton in Lake Shala. Samples were taken at monthly intervals from January to December 2018. A total of 32 zooplankton species were identified, of which 23 species were Rotifers and followed by Copepoda (6) and Cladocera (3). The community was dominated by Brachionus dimidiatus, B. dimidiatus inermis, B. plicatilis and B. calyciflorus. Their abundance exhibit significant seasonal and vertical variations (ANOVA, p < 0.05). Zooplankton community structure of Lake Shala has changed over time, benthic harpacticoids (Nitocra lacustris) have become replaced by small-bodied and generalist-feeder rotifers (genus Brachionus). Such ecological shifts may be driven mainly by partial atelomixis in this deep tropical lake. Overall, zooplankton species distribution and abundance are influenced by various environmental factors such as conductivity, salinity, Chl-a and nutrients. This study concludes that the composition and community structure of zooplankton in Lake Shala has changed over time with the change in physicochemical and internal limnological dynamics (atelomix).

Keywords:

• Brachionus
• ecological shift
• Nitocra lacustris
• rotifers
• vertical variation

1. Introduction

Zooplankton communities play a significant ecological role in aquatic ecosystems (Fernandes and Ramaiah 2019), being a critical link between phytoplankton and higher consumers and they contribute to element recycling and energy transfers (Burian 2010; Helenius 2015). The density and diversity of zooplankton are greatly influenced by different biotic and abiotic factors like physicochemical parameters of water, availability of phytoplankton, lake morphology, and anthropogenic changes in lakes and watersheds (Okoth et al. 2011; Waya et al. 2014; Diego et al. 2015; Merhoon et al. 2017), which alter characteristics and ecology of the lake (Fernández-Álamo and Färber-Lorda 2006; Helenius 2015). Therefore, understanding their dynamics and the factors that affect their community, and their critical linkages with other ecosystem elements, remains an important focus in aquatic ecology to optimizing resource use and enhancing sustainable lake ecosystem management (Joseph and Yamakanamardi 2011; Savitha and Yamakanamardi 2012).

Soda lakes are variable in their physicochemical features and prone to hydrological influences since they are often in semi-arid regions (Oduor and Schagerl 2007; Schagerl et al. 2015; Oduor and Kotut 2016). These factors control the zooplankton diversity and community structure (Burian 2010; Okoth et al. 2011). The zooplankton community structure of saline-alkaline lakes is related to the salinity gradient (Gülle et al. 2010; Anton-Pardo and Armengol 2012; Diego et al. 2015). In addition, saline-alkaline lake zooplankton is strongly influenced by hydrodynamics, while seasonality impacts community composition (Isumbisho et al. 2006; Okoth et al. 2011) or other factors, mainly nutrients, water depth, temperature, dissolved oxygen and food availability can be reasons for vertical and horizontal variations in species composition (Waya et al. 2014; Voutilainen et al. 2016).

Despite the importance of zooplankton in aquatic ecosystems, understanding of their dynamics and the key factors that affect their community structure have so far been unexplained well in alkaline–saline lakes, especially Ethiopian saline-alkaline lakes have received less attention than other kinds of water systems. The only published sets of data on zooplankton of Lake Shala are those of Green and Mengestou (1991), Defaye (1988) and Mengistou (2016), whose investigations involved the analysis of single occasion sampling. Lake Shala offers a good model to investigate the role of partial atelomxis in driving plankton dynamics in the surface mixing layers of such deep tropical lakes, like Lake Zirahuén, Mexico (Tavera and Martı’nez-Almeida 2005), Dom Helvécio and Carioca lakes, Brazil (Souza et al. 2008; Barbosa et al. 2011, 2013). However, few observations have been done on zooplankton dynamics in deep highland lakes in Ethiopia, especially in relation to seasonal mixing patterns (Fetahi et al. 2011; Degefu and Schagerl 2015).

Lake Shala has been facing ecological degradation due to rapidly-increasing human populations and development activities in the basin during the past ten years (Zinabu et al. 2002; Ayenew and Legesse 2007). Furthermore, there is a future plan to expand the Abijata Soda Ash production factory to Lake Shala which may impact the structure of biological communities. Therefore, the present study had 3 major objectives a) prepare a checklist of the zooplankton community which could serve as baseline data to monitor future changes in the lake; b) determine the major drivers of zooplankton dynamics in the lake; c) assess long-term trends in zooplankton composition and biomass in relation to internal and external impacts in the lake (atelomixis or catchment impacts). The study may also contribute towards a better understanding of how climate change and human impacts affect zooplankton dynamics in deep tropical lakes.

2. Materials and methods

2.1. Description of the study area

Lake Shala lies between 7°24′-7°33’N and 38°23′-38°39’E at altitudes of approximately 1558 m within the Abijata-Shala Lakes National Park (Figure 1). The lake is an endorheic and volcano-tectonic lake (Le Turdu et al. 1999). Lake Shala is the deepest among the Ethiopian Rift valley lakes (maximum 266 m), with a surface area of around 329 km² (Von Damm and Edmond 1984; Baxter 2002). This is an inhospitable place, with caustic water concentrated by the high evaporation rate (Ayenew and Legesse 2007). Two rivers (Adabat and Gidu) (Baumann et al. 1975; Baxter 2002) and a number of hot springs supply water to the lake.

Figure 1. Map of Lake Shala showing the study sites/sampling points. (Abbreviation: SPS: Shala Park Shore; SPO: Shala Park Open; SGS: Shala Gike Shore; SGO: Shala Gike Open).

Lake Shala is characterized by a high pH, conductivity, and a high phosphate content, but with very low nitrogen levels (Kebede et al. 1994; Ogato 2015). Despite this hostile nature of the environment, Lake Shala supports phytoplankton and zooplankton, dominated by diatoms and rotifers, respectively (Kebede et al. 1994; Mengistou 2016)). The benthic macroinvertebrate community of the lake comprises Tubificidae, Ostracoda and Chironomidae (Tudorancea and Harrison 1988). Lake Shala and its islands support a large population of Lesser Flamingos (Phoeniconaias minor Geoffroy), Greater White Pelican, Cormorants, Storks and Phalacrocorax carbo (Stephenson 1978; Estifanos 2008; Almaw 2012). The cichlid Oreochromis niloticus and small-sized fish, Aplocheilichthys sp., being reported for the lake (Golubtsov et al. 2002).

2.2. Physicochemical parameters

Physicochemical parameter data were collected together with the zooplankton samples from January to December 2018. The measurements of water temperature, pH, salinity, dissolved oxygen (DO), and electrical conductivity (EC) were conducted in situ with a portable HQ40D Multimeter. For nutrient analysis, composite water samples (by compositing samples from different depths) were taken from each sampling site in triplicate and transported and analyzed spectrophotometrically in the Limnological Laboratory of Addis Ababa University. NO₃-N was analyzed with sodium salicylate method, NH₃-N with indo-phenol blue (APHA 1995), and soluble reactive phosphate (PO₄-P) with ascorbic acid method (APHA, 2000). NO₂-N concentration was determined by the reaction between sulfanilamide and N-naphthyl-(1)-ethylenediamine dihydrochloride (APHA, 1995). Total phosphorus (TP) from unfiltered samples and silica (SiO₂-Si) were determined using persulfate digestion and Molybdosilicate Method, respectively (APHA, 1999). Total alkalinity (TA) was determined by titration with 1 N HCl to pH 4.5 using a mixed indicator (bromocresol green-methyl red) within a few hours of sample collection according to Wetzel and Likens (2000).

2.3. Zooplankton and other biological data sampling

Zooplankton collected at monthly intervals from January to December 2018 from four sampling stations (Figure 1). Zooplankton samples were obtained by vertically hauled from 30 m to the surface from two open water stations and 10 m from two shore sites through a 30-μm plankton net. A vertical bottle sampler (van Dorn, 2 L capacity) was also used for the determination of vertical distribution and abundance of zooplankton at the water surface (0 m), 5, 10, 15, 20, and 30 m depth of open water sampling station and water temperature, electric conductivity and dissolved oxygen were measured at each depth simultaneously with zooplankton sampling. The zooplankton samples were preserved in 4% formalin and brought to Addis Ababa University, Limnology Laboratory. Zooplankton was identified as the lowest taxonomic unit according to these taxonomic keys: Copepods (Defaye 1988), Cladocerans (Korinek 1999) and Rotifers (Fernando 2002). Counting was carried out using a plankton counting chamber under an inverted microscope. Total counts were made and individual densities were expressed as numbers per liter. The abundance of phytoplankton was calculated using the equation in Hötzel and Croome (1999) and Phytoplanktonic Chlorophyll-a (Chl-a) concentrations were estimated from water samples collected from the same place where the in situ variables were measured. Chl a was extracted using 90% acetone by following the procedure of Talling and Driver (1963). Fish pressure.on zooplankton aboundance and diversity was determined by fish stomach content analysis (Hyslop 1980). Macro-invertebrates were collected based on the method outlined in Ontario Benthos Biomonitoring Network Protocol Manual (Jones et al. 2007).

2.4. Data analyses

Spatial and temporal variability of zooplankton community between sites, and among seasons, were analyzed by non-parametric Kruskall–Wallis ANOVA (Kruskal and Wallis 1952). The association between the distribution of zooplankton species, abiotic and biotic factors and sampling stations was evaluated with canonical multivariate analysis using CANOCO windows 4.5 version Software (Ter Braak and Smilauer 2002). Detrended correspondence analysis (DCA) was employed to check the response of the data, and the length gradient was 0.195. Therefore, redundancy analysis (RDA) was used to determine the relationship between zooplankton species distribution and environmental parameters.

3. Results

3.1. Spatial, seasonal and trends in physicochemical parameters

The seasonal and spatial environmental variables of Lake Shala are presented in Table 1 and these variables showed evident spatial and seasonal variations (ANOVA: p < 0.05), except DO, which showed only seasonal variation in Lake Shala (Tables 1 and 2). pH (10.08), Alkalinity (291.1 meq/L), NO₃-N (0.17 μg/L), NH₃-N (108.78 μg/L), SiO₂ (1.09 mg/L), TP (2.03 mg/L) and SRP (1.25 mg/L) were significant at Shala Gike Shore (SGS).

Table 1. Spatial variation of physical environmental variables of the studied stations of Lake Shala (SPS: Shala Park Shore; SPO: Shala Park Open; SGS: Shala Gike Shore; SGO: Shala Gike Open; Temp: Temperature; DO: Dissolved Oxygen; EC: Electrical Conductivity; Alka: Alkalinity).

	pH	Temp	EC	DO	Sal	Alka	NO3-N	NH3-N	SiO2	TP	SRP
SPS	9.94 ± 0.27^a	25.1 ± 1.2^b	31.5 ± 3.5^a	7.3 ± 1.1^a	19.6 ± 2.4^a	275.1 ± 28.3^a	0.16 ± 0.04^ab	100.3 ± 26.6^a	0.99 ± 0.4^a	1.8 ± 0.4^a	1.2 ± 0.5^a
SPO	9.96 ± 0.14^ab	24.6 ± 1.5^a	32.9 ± 2.8^b	7.6 ± 0.6^a	20.6 ± 2^b	252.9 ± 10.6^b	0.14 ± 0.02^a	71.4 ± 14.7^b	0.67 ± 0.2^b	1.3 ± 0.4^b	0.9 ± 0.4^b
SGS	10.08 ± 0.15^c	24.7 ± 0.8^ab	30.4 ± 1.1^a	7.4 ± 1.1^a	18.7 ± 1.3^a	291.1 ± 16.1^c	0.17 ± 0.05^b	108.8 ± 36.9^a	1.09 ± 0.4^a	2.0 ± 0.4^c	1.3 ± 0.5^a
SGO	10.06 ± 0.16^bc	24.2 ± 0.8^a	31.4 ± 1^a	7.4 ± 1^a	19.5 ± 0.7^a	268.1 ± 14.2^a	0.15 ± 0.03^a	74.9 ± 16.6^b	0.72 ± 0.3^b	1.5 ± 0.5^b	0.8 ± 0.4^b
Mean	10.01 ± 0.19	24.6 ± 1.1	31.5 ± 2.6	7.4 ± 0.9	19.6 ± 1.8	271.8 ± 22.9	0.15 ± 0.04	88.8 ± 29.8	0.87 ± 0.36	1.7 ± 0.5	1.0 ± 0.5

Note. Values with different letters (a, b, c) within a column are significantly different at p < 0.05 level (Tukey test is applied).

Table 2. Seasonal variation of physical environmental variables of Lake Shala (Temp: Temperature; DO: Dissolved Oxygen; EC: Electrical Conductivity; Alka: Alkalinity).

	pH	Temp	DO	EC	Sal	Alka	NO3–N	NH3–N	SiO2	TP	SRP
Dry	9.97 ± 0.2^a	24.2 ± 0.9^a	8.1 ± 0.8^a	34 ± 3.4^c	21.3 ± 2.4^c	268.4 ± 20.2^a	0.12 ± 0.02^a	64.8 ± 14.8^a	0.6 ± 0.3^a	1.3 ± 0.4^a	0.5 ± 0.2^a
Pre rainy	9.97 ± 0.2^a	25.2 ± 1.3^b	7.3 ± 1.3^b	30 ± 2^a	18.4 ± 1.6^a	268.2 ± 20.7^a	0.15 ± 0.02^b	92.5 ± 31.1^b	1 ± 0.4^b	1.8 ± 0.5^b	1.1 ± 0.4^b
Rainy	10.1 ± 0.1^b	24.5 ± 1.1^a	6.9 ± 0.7^b	30.8 ± 1^ab	19.1 ± 0.6^ab	280.5 ± 18.1^b	0.19 ± 0.04^c	113.7 ± 31.9^c	1.2 ± 0.3^c	2.1 ± 0.4^c	1.5 ± 0.2^c
Post rainy	9.99 ± 0.3^a	24.7 ± 0.7^a	7.3 ± 0.4^b	31.4 ± 0.9^b	19.5 ± 0.6^b	270.2 ± 29.6^ab	0.16 ± 0.01^b	84.3 ± 12.4^b	0.7 ± 0.1^b	1.4 ± 0.3^a	1.0 ± 0.4^b

Mean concentrations of pH (10.1), Alkalinity (380.5 meq/L), NO₃-N (0.19 μg/L), NH₃-N (113.69 μg/L), SiO₂ (1.16 mg/L), TP(2.14 mg/L) and SRP (1.45 mg/L) were recorded during rainy season, whereas DO (8.12 mg/L), EC (33.99 mS/cm) and salinity (21.34 g/L) were recorded during the dry season and were significantly different.

Comparing the present physicochemical variables value to studies made during 1961-2012, a long-term increasing trend in pH, Alkalinity, Salinity, EC, NO₃-N, NH₃-N, SRP and TP, and a decreasing trend in SiO₂ were observed (Table 3). The mean electrical conductivity was generally higher in the present study compared to data from the 1960s (Table 3) and nearly comparable to the mean value (33000 μS/cm) reported by Talling and Talling (1965) (Table 3). The current mean value of salinity in Lake Shala (19.6 g L⁻¹) is comparable to the value (19.582 g L⁻¹) reported by Talling and Talling (1965), but higher than that reported by Baumann et al. (1975) (16.8 mg L⁻¹); 18.1 mg L⁻¹ Kebede et al. (1994) and 15.7 mg L⁻¹ (Talling and Talling 1965).

Table 3. Some previous and current physicochemical factors of Lake Shala, and trends over 50 years back from the present study.

References	Sampling time	pH	EC	Salinity	Alkalinity	SRP	SiO2	NO3-N	NH3-N
Talling and Talling (1965)	May 1961		33,000	19.582	200		130
Wood and Talling (1988)	Jan 1964				210
Wood and Talling (1988)	Mar 1964	9.9	21,940	21.5	212	760	112
Wood and Talling (1988)	Oct 1966		24,640		216
Baumann et al. (1975)	May 1971	10		16.8	188
Von Damm and Edmond (1984)	Jan 1976				218.8
Kebede et al. (1994)	March-May 1991	9.65	21,940	18.1	218	809	56	ND	4.3
Zinabu et al. (2002)	1990-2000	9.7	22,863		217.3	727	49.1
Zinabu (2002)	Wet and dry seasons (1990-2000)	9.72	22,757		216.71	613.85	49.5	0.85	4.65
Ogato (2015)	Feb 2012 to Jan 2013	9.96	25,800	15.7	224	961	0.9	ND	42
Present study		10.01	31,540	19.6	271.8	1020	0.87	0.154	88.82

The concentration of soluble reactive phosphorus (SRP) appears to have decreased from 1964 to 2000 in Lake Shala except in 1991 (809 μg L⁻¹) a record by Kebede et al. (1994), however highest record was reported during the recent year sampling period by Ogato (2015) (961 μg L⁻¹) and 1020 μg L⁻¹ in the present study. Concentrations of silicate decreased markedly in Lake Shala. Compared to the present data, the high silicate value (130 mg L⁻¹) were recorded in the 1960s by Talling and Talling (1965). Ogato (2015) measured the lowest concentration in 2012 comparable to the current result (Table 3). The mean concentration of NH₃-N increased considerably when compared with some results reported by Kebede et al. (1994) (4.3 μg L⁻¹), Zinabu (2002) (4.65 μg L⁻¹), Ogato (2015) (42 μg L⁻¹).

3.2. Zooplankton composition and abundance

A total of 32 zooplankton taxa were recorded in Lake Shala across all sites and in the studied periods (Table 4). Of these species composition of zooplankton of Lake Shala, the rotifers had the highest species richness of the total zooplankton (23 taxa), of the 23 rotifers identified, 10 species belonged to the genus Brachionus including B. budapestinesis, B. calyciflorus, B. dimidiatus, B. dimidiatus inermis, B. falcatus, B. plicatilis, B. quadridentatus, B. rubens, and B. urceolaris. Other groups were represented by Copepoda (6 taxa) and Cladocera (3 taxa) (Table 4).

Table 4. Zooplankton taxa recorded in Lake Shala from January- December 2018.1.

Arguloida (Copepoda)	Cyclopoida (Copepoda)	Harpacticoida (Copepoda)	Cladocera	Rotifera
Argulus africanus	Afrocyclops gibsoni	Nitocra lacustris	Diaphanosoma excisum	Anuraeopsis fissa
	Eucyclops serrulatus		Moina belli	Anuraeopsis navicula
	Thermocyclops ethiopiensis		Moina micrura	Brachionus angularis
	Thermocyclops oblongatus			Brachionus budapestinesis
				Brachionus calyciflorus
				Brachionus dimidiatus
				Brachionus dimidiatus inermis
				Brachionus falcatus
				Brachionus plicatilis
				Brachionus quadridentatus
				Brachionus rubens
				Brachionus urceolaris
				Filinia pejleri
				Hexarthra jenkina
				Keratella tropica
				Lecane hastate
				Lecane bulla
				Lecane luna
				Polyarthra spp.
				Trichocerca tetractis
				Trichocerca spp.
				Testudinella patina trilobata

The mean abundance of zooplankton ranged from 830.4 to 1538.6 ind.L⁻¹, with the maximum values in August and the minimum in January (Figure 2). Among all the zooplankton samples, rotifers accounted for the most considerable fraction of zooplankton density and contributed 86.11% of the total zooplankton density (Figure 3). Among the rotifers, Brachionus dimidiatus, B. dimidiatus inermis, B. plicatilis and B. calyciflorus dominated during all sampling periods. The highest cell density of B. dimidiatus occurred during the dry (January - March) and post rainy months (September -December). The cell density of B. dimidiatus inermis, however, increased during the pre-rainy months (April - May). Similar patterns of B. plicatilis and B. calyciflorus were observed, with increasing abundance during the rainy months (July - September) (Figure 4).

Figure 2. Monthly pattern of total zooplankton abundance (ind. L⁻¹) in Lake Shala from January to December 2018.

Figure 3. Monthly distribution of zooplankton groups abundance (ind. L⁻¹) in Lake Shala from January to December 2018.

Figure 4. Temporal variation of the four dominant rotifer species in Lake Shala from January to December 2018.

3.3. Spatial and seasonal distribution of zooplankton

The spatial and seasonal variations in the abundance of zooplankton were computed for all the sampling sites and seasons (Tables 5 and 6). On average SPS site had the highest zooplankton mean abundance of 1198.5 ± 296.3 ind.L⁻¹ followed by SPO (1145.0 ± 214.8 ind.L⁻¹), SGO (1109.6 ± 257.4 ind.L⁻¹) and SGS (1076.6 ± 360.2 ind.L⁻¹) (Table 5). However, there was no significant difference in spatial zooplankton abundance among the four stations (F = 0.395, p = 0.757).

Table 5. Spatial and total abundance variations in the zooplankton community.

	Rotifer	Nauplii	Copepoda	Cladocera	Total Abundance
SPS	1020.9 ± 242.9	145.8 ± 84.4	26.4 ± 4.8	5.3 ± 2.0	1198.5 ± 296.3
SPO	964.6 ± 185.9	148.4 ± 99.4	27.2 ± 5.5	4.8 ± 1.7	1145.0 ± 214.8
SGS	1019.7 ± 224.7	87.5 ± 33.4	27.6 ± 5.7	4.8 ± 1.7	1139.7 ± 222.8
SGO	958.4 ± 249.4	119.0 ± 77.8	27.6 ± 5.7	4.6 ± 1.5	1109.6 ± 257.4
Mean	990.9 ± 221.7	125.2 ± 79.2	27.2 ± 5.1	4.9 ± 1.7	1148.2 ± 243.9

Table 6. Seasonal variations in the abundance (ind.L⁻¹) of the zooplankton group.

	Rotifer	Nauplii	Copepoda	Cladocera	Total Abundance
Dry	795.5 ± 161.6^a	144.3 ± 96.1^a	24.8 ± 4.1^a	3.1 ± 0.5^a	967.8 ± 209.0^a
Pre rain	1021.7 ± 128.1^b	66.4 ± 36.3^b	26.2 ± 4.7^a	4.1 ± 1.0^b	1118..4 ± 142.7^a
Rain	1256.8 ± 97.4^c	138.2 ± 85.5^a	33.5 ± 1.1^b	7.1 ± 0.7^c	1435.5 ± 159.5^b
Post rain	889.8 ± 164.5^ab	151.8 ± 61.9^a	24.2 ± 3.5^a	5.3 ± 0.8^d	1071.1 ± 178.2^a
Mean	990.9 ± 221.7	125.2 ± 79.2	27.2 ± 5.1	4.9 ± 1.7	1132.4 ± 281.6

The seasonal maximal abundance of zooplankton was observed during the rainy season as 1438.1 ± 159.5 ind.L⁻¹, and minimum abundance during the dry seasons as 967.8 ± 209.0 ind.L⁻¹, thus seasonal fluctuation in the zooplankton community was important (p < 0.05) (Table 6). Rotifer, Nauplii, Copepoda and Cladocera abundance (ind.L⁻¹) exhibited significant seasonal pattern, showing a maximum during rainy seasons for Rotifers, Copepoda and Cladocera while Nauplii during post rainy seasons (Table 6).

3.4. Vertical distribution of zooplankton

The vertical distribution and percentage contribution of the zooplankton group in Lake Shala are shown in Figure 5. The vertical distribution of zooplankton demonstrated a significant correlation between zooplankton groups and water depth. The mean zooplankton abundance was greatest in the top 10 m of the water column, having its maximum in the surface (1,357. 7 ind.L⁻¹). However, a size-structured habitat utilization by zooplankton populations in the lake varied in depth. The average density of rotifer and nauplii decreased with increasing depth, it attained the highest averages of 712.7 ind.L⁻¹ and 504.1 ind.L⁻¹ at the surface and it decreased to 1.6 ind.L⁻¹ and 39.6 ind. L⁻¹ at 15–20 m, respectively (Figure 5). The greatest contrast in the distribution and percentage contribution was observed for the copepod and cladoceran, which were concentrated in the mid and deeper layer of the water, respectively (Figure 5). Vertical electrical conductivity, temperature, and dissolved oxygen profile of the lake were determined in the same period (Figure 6). The vertical distribution patterns of zooplankton were remarkably similar to that of electrical conductivity, temperature, and dissolved oxygen curves (Figures 5 and 6).

Figure 5. Mean vertical distributions of zooplankton in Lake Shala from January to December 2018.

Figure 6. Dissolved oxygen (DO), Electrical conductivity (EC) and Water temperature profile of Lake Shala from January to December 2018.

3.5. Zooplankton species and environmental variables

The Redundancy Analysis (RDA) summarized 94.5% of the total variance in species data on the first two axes (axis 1, 73.5%; axis 2, 20.9%). Some zooplankton species were associated with environmental variables: (i) Brachionus angularis, B. urceolaris, B. plicatilis, Thermocyclops ethiopiensis, and Diaphanosoma excisum, highlighted on the right-hand side were associated with high levels of SRP, NH₃-N, NO₃-N, TP and macroinvertebrate, and negatively correlated with salinity and electrical conductivity; (ii) Brachionus dimidiatus, B. rubens, B. quadridentatus and T. oblongatus were positively correlated with Alkalinity, Chl a, phytoplankton abundance, pH and fish pressure; (iii) Moina micrura, M. belli, Afrocyclops gibsoni, Eucyclops serrulatus and nauplii shown on the left-hand side were associated with salinity, electrical conductivity and dissolved oxygen; (iv) Brachionus dimidiatus inermis and Brachionus calyciflorus were positively correlated with temperature. In addition, most species shown in the upper half of the tri-plot were negatively correlated with salinity and EC (Figure 7).

Figure 7. Redundancy analysis (RDA) triplot of zooplankton species in relation to water physicochemical properties and sites (E. serru. - Eucyclops serrulatus, A. gibso. -Afrocyclops gibsoni, M. micru. - Moina micrura, M. belli. - Moina belli, B. calyc. - Brachionus calyciflorus, B.d. in. - Brachionus dimidiatus inermis, D. excis.-Diaphanosoma excisum, B. angul.- Brachionus angularis, T. ethio. - Thermocyclops ethiopiensis, B. plica. - Brachionus plicatilis, B. urceol. - Brachionus urceolaris, T. oblong. - Thermocyclops oblongatus, B. quadr.- Brachionus quadridentatus, B. dimid. - Brachionus dimidiatus, B. ruben. - Brachionus rubens, SPS-Shala Park Shore; SPO-Shala Park Open; SGS-Shala Gike Shore; SGO-Shala Gike Open; Alka-Alkalinity; Chl a - Chlorophyll; Phyto Ab - Phytoplankton Abundance).

4. Discussion

4.1. Composition and distribution of zooplankton

Zooplankton composition of Lake Shala exhibited typically tropical saline inland water euryhaline biota forms (Derry et al. 2003; Burian 2010; Okoth et al. 2011; Ong’ondo et al. 2013), wherein few tolerant and adapted species attained high population densities, as previously reported for this lake (Green and Mengestou 1991). In the present study, euryhaline rotifers and crustaceans such as Brachionus dimidiates, B. plicatilis, B. calyciflorus, Afrocyclops gibsoni, Eucyclops serrulatus, Thermocyclops oblongata and T. ethiopiensis inhabited the lake. These rotifers and crustacean species are cosmopolitan and broadly distributed given their broad tolerance to salinity (Derry et al. 2003; Echaniz et al. 2006; Okoth et al. 2011).

In the present study, small-sized rotifers (B. dimidiatus, B. dimidiatus inermis, B. plicatilis and B. calyciflorus) were numerically dominant in the zooplankton population with a significant difference. Similarly, several studies reported the dominance of rotifers in saline-alkaline lakes such as Lake Nakuru, Bogoria, and Abijata (Wodajo and Belay 1984; Green and Mengestou 1991; Burian 2010; Okoth et al. 2011). The leading factors to explain this dominance by rotifers are attributed to their fast reproductive rates and developmental stages (Yin et al. 2016), unspecialized feeding (Joseph and Yamakanamardi 2011) and low abundance and diversity of cladocerans (Anton-Pardo and Armengol 2012) in saline lakes promote the development of rotifers. Also rotifers filter small-sized diatoms more efficiently that large-sized Cladocera and Copepoda in the mixing layer in Lake Shala.

The species richness of cladocerans was low and they occurred rarely, according to Afonina and Tashlykova (2018) this group is typical of freshwater, and most species do not tolerate high salinity and electrical conductivity. In East Africa Soda Lakes, species richness and composition are correlated positively with salinity (Mengestou, Mengistou 2016). The species richness of Cladocerans decreases with increasing levels of salinity and electrical conductivity of the water body (Diego et al. 2015). High salinity and conductivity in Lake Shala also restricted the diversity of Cladocerans species. This agrees with the observations of Ong’ondo et al. (2013) and Wodajo and Belay (1984), who reported low or no record of Cladocera diversities in soda lakes due to high conductivity and salinity, which limits osmoregulation capacity, reduce reproductive and survival rates of Cladocerans (Tavsanoglu et al. 2015; Afonina and Tashlykova 2018).

Even though significant differences in environmental factors such as physicochemical parameters were recorded in four sampling stations, the rotifer, nauplii, copepoda and cladoceran abundance did not exhibit significant differences among sampling stations. The relative zooplankton abundance homogeneity among sites of Lake Shala follows the findings of Nandini et al. (2008) in the Valle de Bravo reservoir. However, Dagne et al. (2008) and Yongo and Outa (2017) found the strongest horizontal gradients in large-bodied zooplankton. In the present study, the observed lack of horizontal or spatial distribution is not a surprise since the lake is dominated by small-bodied taxa and juvenile ontogenetic stages which have low susceptibility to predation and coupled with the low locomotory ability adequately explains a lack of spatial variations in distribution. Although the physical structure of Lake Shala is distinguished by high wind mixing in the afternoon (partial atelomixis), which may have a significant contribution to the distinct lack of horizontal variations in zooplankton distribution. Souza et al. (2008) and Degefu and Schagerl (2015) also mentioned the key role of atelomixis as a critical factor in the tropical lakes for the plankton community structure.

Several studies on other saline-alkaline lakes documented seasonal variations in zooplankton abundance (Wodajo and Belay 1984; Derry et al. 2003; Gülle et al. 2010; Okoth et al. 2011; Afonina and Tashlykova 2018). These studies were on deep, oligotrophic lakes such as Lake Burdur (Gülle et al. 2010), and shallow alkaline–saline lakes of Ethiopia (Lake Abijata) and Kenya (Lake Nakuru and Bogoria) (Wodajo and Belay 1984; Burian 2010; Okoth et al. 2011). Those authors observed a clear seasonality pattern in the abundance of different zooplankton taxa. The present study also exhibited seasonal variations with the increasing trend during the rainy season, similar to the observed rise of phytoplankton density and biomass (Wagaw et al. 2021), which resulted in greater abundance and possibly quality of edible resources (Nandini et al. 2008; Arora and Mehra 2009; Joseph and Yamakanamardi 2011). Okoth et al. (2011) and Gülle et al. (2010) also observed the highest zooplankton abundance during periods of long rains in saline-alkaline lakes of Nakuru and Burdur.

The vertical distribution of zooplankton has been studied in various deep Lakes (e.g. Gülle et al. 2010; Degefu and Schagerl 2015). Usually, zooplankton shows distinct vertical distribution rather than being homogeneously distributed through the water column. During the present study, the vertical distribution of different zooplankton taxa was concentrated in various zones, nevertheless, the most concentrated region was between 0-15 m (p < 0.05). Most zooplankton density that exists in the upper 15 meters (euphotic layer) of Lake Shala may be due to the abundance of phytoplankton. Gutkowska et al. (2018), discussed and indicated a significant direct correlation between phytoplankton and zooplankton population density in Vistula Lagoon and Lake Łebsko. According to this author, the phytoplankton abundance and biomass showed a pronounced significant occurrence in the euphotic zone than in the non-euphotic layer, which resulted in a dense zooplankton population. Moreover, Lake Shala’s surface water temperatures are tropical, and the minima continue to be high enough to support continued zooplankton production (Burian 2010; Okoth et al. 2011).

However, during persistent partial atelomictic mixing, zooplankton favor the mixing epilimnion (mixolinion), indicating that food conditions are ideal in this water column. This is likely because of frequent diurnal partial atelomictic mixing, which suspends food particles like phytoplankton, protozoans, and bacteria freely in the upper zone of the water column and permits a rapidly growing zooplankton population in Lake Shala, as occurs in other deep tropical lakes after atelomixis (Ciros‐Pérez et al. 2015; Degefu and Schagerl 2015). Rotifers and nauplii in particular may experience rapid population peak if food resource increase due to frequent mixing due to atelomixis and contributing for the unpredictable rises in population abundance (Ciros-Pérez et al. 2001). Ortega-Mayagoitia et al. (2011) also verified that rotifers as opportunistic species and able to take advantage over increasing in food availability for their growth (especially Brachionus). Another important factors that explain variations in zooplankton abundance could be light-related tolerance factors, oxygen saturation, competition and predation (Williamson et al. 2020). According to Ciros‐Pérez et al. (2015), copepods and cladocera migrated to a zone with very low levels of visible light in deep tropical oligotrophic lake.

The most notable occurrence of rotifers and nauplius was detected as the most dominant species in 0 - 10 m depth throughout the year and their density decreased with increasing depth (Figure 5). These differences were significant for rotifers (F = 17.65, p = 0.00, ANOVA) and nauplius (F = 28.898, p = 0.00, ANOVA) between the various sampling depth. Rotifera attained its highest density year-round due to the flourishing of Brachionus species which can tolerate a high temperature, salinity, and electrical conductivity. Moreover, low phytoplankton biovolume, species composition, and anoxic conditions caused by a steady drop in temperature with depth are related to the decreased density of rotifers and nauplii with an increase in depth (Gülle et al. 2010). In this study, the phytoplankton composition and Chl a biomass in relation to depth were not analyzed, but a decline and variance distribution of phytoplankton and Chl a biomass occur in relation to depth (Żurek and Bucka 1994; Nandini et al. 2008; Voutilainen et al. 2016).This finding is concordant with Isumbisho et al. (2006) in Lake Kivu and Gülle et al. (2010) in Lake Burdu, which reported rotifers and nauplii larvae inhabited the surface layer (1-10) and the deeper by Copepods and Cladoceran.

In this study, the vertical distribution patterns of zooplankton were similar to that of electrical conductivity, temperature and dissolved oxygen curves (Figures 5 and 6). Copepods and Cladocerans are known to vertically migrate even in the lower oxygen condition (Dimante-Deimantovica et al. 2012; Voutilainen et al. 2016). In this study, the abundance of Copepods (F = 19.899, p = 0.00, ANOVA) and Cladocerans (F = 11.908, p = 0.00, ANOVA) were different among the sampling depth and the deepest region had the highest densities. This is supported by findings from various authors (Isumbisho et al. 2006; Gülle et al. 2010) who reported depth and electrical conductivity as a key factor governing changes in zooplankton abundance. Also, Copepods and Cladocerans are in the depths because of their inability to tolerate the light intensity on the surface compared to the depths (Gülle et al. 2010; Merhoon et al. 2017) and this can also explain the increase in the total density in the depths compared to the surface.

4.2. Driving factors of zooplankton abundance

By comparing the zooplankton community structure with environmental factors, the zooplankton community varied significantly as biotic and abiotic factors changed between different sampling stations. Chl a, phytoplankton abundance, pH, Alkalinity, and TP, reached the highest values at Shala Gike Station (SGS), where we find abundant B. dimidiatus, B. rubens, B. quadridentatus and T. oblongatus species. Temperature, NH₃–N and SRP were also positively correlated with zooplankton abundance, which contributes to the high abundance of D. excisum T. ethiopiensis B. urceolaris, B. budapestinesis and B. plicatilis. Several earlier studies have also confirmed that abiotic and biotic factors such as temperature, nutrients (N and P), SiO₂, pH, ionic concentrations and predation influence, directly or indirectly, the distribution and abundance of zooplankton species (Derry et al. 2003; Waya et al. 2014; Tavsanoglu et al. 2015).

The present study showed that the zooplankton community structure is generally influenced by EC and salinity, and most of the species were found against them. Gülle et al. (2010) and Afonina and Tashlykova (2018), also pointed electrical conductivity (salinity) as one of the several factors responsible for the structure of zooplankton communities in saline-alkaline lakes. Its impact can be strongly modified by other factors, and there are also indirect mechanisms by which it may exert its effects e.g. by imposing osmoregulatory stress, influence the availability of feed, or inducing migrations of organisms trying to evade high or low salinity and impact on the abundance of the zooplankton population (Perumal et al. 2009.; Afonina and Tashlykova 2018; Gutkowska et al. 2018).

4.3. Long-term trends of zooplankton changes in deep tropical Lake Shala

The zooplankton taxa encountered in Lake Shala during this study differed remarkably from those previously reported for this lake (Defaye 1988; Green and Mengestou 1991; Mengistou 2016), where between 1 and 9 zooplankton species were recorded. During the previous studies and Mengistou (2016) in his review, of 12 identified species, 9 belonged to Rotifera; 2 Copepoda and 1 Cladocera species were reported. Partly similar results was observed in the present study; of 32 identified species, 23 belonged to Rotifera, 6 Copepoda and 3 Cladocera species was identified. The results of our study revealed that the composition of the species in zooplankton varied considerably, there was a significant increase in the number of species. Because the plankton mesh size net in this study (30 µm) differed from those of earlier studies (64 and 55 µm), the difference in sample collection methods, and inclusion of open water system and deep species, which might bring variation in the zooplankton community composition.

The zooplankton community structure of Lake Shala has also changed over time Tudorancea and Taylor (2002) and Mengistou (2016) reported benthic harpacticoids (Nitocra lacustris) as dominant species from Lakes Shala. However, in the current study, Nitocra lacustris were replaced by small-bodied and generalist-feeders rotifers (genus Brachionus) as the dominant taxa within all sampling sites and seasons. The notable absence of dominance of the benthic harpacticoids in Lake Shala might indicate the presence of ecological or major habitat shifts in the lake over the years. Also, the shift might be due to several reasons including the changes in the water chemistry of the lake (Table 4). Various studies in East African soda lakes revealed that plankton community structure changes in relation with water chemistry (Schagerl and Oduor 2008; Ogato and Kifle 2017). Zooplankton community shift may also associated with ‘atelomixis’ changes brought about by unusual mixing behavior in deep, tropical lakes (Fetahi et al. 2011; Degefu and Schagerl 2015) (Table 7).

Table 7. List and number of zooplankton taxa reported from Lake Shala.

Taxon	Defaye (1988)	Green and Mengestou (1991)	Mengistou (2016)
Cyclopoida (Copepoda)	Thermocyclops ethiopiensis
Harpacticoida (Copepoda)			Nitocra lacustris
Cladocera			Moina belli
Rotifera		Brachionus dimidiatus B. plicatilis Hexarthra jenkinae Lacinularia elliptica	Brachionus dimidiatus B. plicatilis B. urceolaris B. angularis Lecane luna L. bulla Trichocerca tetractis
	N = 1	N = 4	N = 9

5. Conclusion

In conclusion, euryhaline rotifers dominance in deep tropical lakes is observed, show variations in seasonal and vertical distribution: rotifers and nauplii concentrate in the epilimnion, while copepods and cladoceran are found predominantly in the deeper zone of the lake. Salinity and electrical conductivity remain the most critical factor influencing zooplankton distribution. A noticeable decrease in density and species richness of zooplankton was noticed at sites with high salinity and electrical conductivity. The community composition was highly changed with time series. Thirty-two taxa were recorded during the study, dominance copepod Nitocra lacustris were replaced by rotifers (genus Brachionus). This long-term succession of zooplankton may be associated with physicochemical changes and unusual mixing behavior (partial atelomixis) in deep tropical lakes. Also in future zooplankton studies in the lake, determination of diel vertical migration patterns and vertical distribution effects of the physicochemical variables, thermal stratification, phytoplankton abundance and chlorophyll a would contribute to further understanding of the deep tropical dynamics of the lake zooplankton.

Acknowledgements

We appreciate the financial support provided by Addis Ababa University through its Graduate Studies Program and Thematic Research Project. We thank Dr. Assefa Wosnie and Dr. Yirga Enawgaw for their help and advice in fieldwork and research techniques. Finally, we thank the Ethiopian Wildlife Conservation Authority (EWCA) and Abijatta-Shala Lakes National Park Administration for granting permission to conduct research in Lake Shala.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The present study was financed by Graduate Studies Program and Water Thematic Research Project of Addis Ababa University.

Notes on contributors

Solomon Wagaw

Dr. Solomon Wagaw graduated from Hawassa University with a first-class biology degree in 2009, and he also earned a master’s degree in limnology, fisheries, and aquatic ecotoxicology there in 2012. His Ph.D. in Aquatic Sciences, Fisheries, and Aquaculture was awarded by Addis Ababa University. He’s written eight publications that were peer-reviewed. Solomon is currently working as lecturer and researcher at Wolkite University.

Seyoum Mengistou

Professor Seyoum Mengistou received his Ph.D. from Waterloo University in Biology (Limnology). He has held a number of positions, including those of director of the zoological natural museum, head of the biology department at Addis Abeba University, editor-in-chief of the Ethiopian Journal of Biological Sciences, founding member of the Ethiopian Fisheries and Aquatic Sciences Association (EFASA), member of the Biological Society of Ethiopia, and member of the Society International of Limnology. Over 20 Ph.D. students have been mentored and supervised by him, and he has published over 95 publications in esteemed national and international scientific journals.

Abebe Getahun

Professor Abebe Getahun earned his BSc in Biology from Addis Abeba University as well as his MSc in Zoology. He then attended a combined program run by the American Museum of Natural History and the City University of New York, USA, to earn his MPhil and PhD in Ecology, Evolution, and Behaviour. He is working on research initiatives while also instructing undergraduate and graduate (MSc and PhD) students. The diversity, conservation, and use of aquatic organisms, particularly fish, are the main areas of his research. Over 90 articles in books and scholarly publications were written by him or with whom he collaborated. He has won numerous funding for study from various organizations. He has held a number of positions in academic administration at Addis Abeba University. He was a founder member of the Biological Society of Ethiopia (BSE) and the Ethiopian Fisheries and Aquatic Sciences Association (EFASA). He is a SINET Editor: Ethiopian Journal of Science and Ethiopian Journal of Biological Sciences.