Spatial distribution of zooplankton in response to ecological dynamics in tropical shallow lake: insight from Lake Malombe, Malawi

Abstract This study assessed zooplankton density and Chl-a amount to understand their spatial and seasonal variation in Lake Malombe. Samples were collected for analysis during the hot, dry season (HDS) and cool, dry, windy season (CDWS). The zooplankton identified were Tropodiaptomus cunningtoni (TC), Mesocyclops aequatorialis aequatorialis (MAA), Thermocyclops neglectus (TN), Bosmina longirostris (BL), Diaphanosoma spp. (DS), and Nauplii. These zooplankton groups belong to copepods, Cyclopidae, Cladocera, Ctenopod, and Rotifers. Chl-a exhibited a substantial seasonal variation, with the highest concentration observed in HDS and lowest in CDWS. Zooplankton such as T. cunningtoni, M. aeq. aequatorialis, T. neglectus, B. longirostris, Diaphanosoma spp, Nauplii, and Rotifer increased during the HDS. The water temperatures (WT), pH, Dissolved Oxygen (DO), and Chl-a positively correlated with all zooplankton densities. Sodium (Na+) and Potassium (K+) ions positively correlated with TC, MAA, DS, Nauplii, and Rotifers. Chloride ion (Cl-) positively correlated with DS, Rotifers, while Carbonate ion (CO3 2-) positively correlated with TC. Bicarbonate (HCO3 -) correlated positively with DS and Rotifers, while TDS correlated negatively with BL, Nauplii, and Rotifers. Nitrite affected zooplankton density negatively, while SRP had a positive effect. The study revealed that physical-chemical variables, some significant ions, and Chl-a are the most critical factors influencing the spatial and seasonal variation of zooplankton density in the lake. These findings demonstrate the interaction between physical-chemical variables, Chl-a, and zooplankton and highlight the significance of ecological understanding of the complex dynamics of the food web system in shallow tropical lakes such as Lake Malombe- under the changing climate.


Introduction
Tropical Inland African Rift Valley lakes are characterized by high carbonate salts concentration and are often described as alkaline lakes (De Cort et al. 2019). These lakes lack connection to the marine environment, and hence some of them are described as freshwater lakes (Butturini et al. 2020). They possess tremendous ecological values to the local communities regarding provisioning, regulatory, supporting, cultural, and aesthetic services (IPBES 2018). These lakes are best known for having the world's richest lacustrine fish fauna and constitute the family of Cichlidae, which provide the supreme example of geographically circumscribed vertebrate Evolution and a unique comparative series of natural laboratories for evolutionary studies (Weyl et al. 2010;Gess and Whitfield 2020). Despite the ecological importance of these lakes, studies focusing on zooplankton species distribution and how they are affected by seasonal dynamics of Chl-a and physical-chemical variables are scarce (Gebrehiwot et al. 2017;Ma et al. 2019). Several ecological factors such as water temperatures (WT), electrical conductivity (EC), nutrient concentration, calcium ions, magnesium ions, and hydrological pattern have been considered as critical factors responsible for shaping the biological composition and distribution of zooplankton in inland lakes (Ha et al. 1998;Teubner 2003;Afonina and Tashlykova 2018;Hamid et al. 2020). In regions where seasonality is very distinct, variation in environmental factors such as WT, pH, and nutrient concentration plays a significant role in zooplankton composition and distribution via the influence of light and nutrients concentrations (Liu et al. 2010;Vidal et al. 2017). Different researchers have extensively studied the zooplankton composition and structure in tropical freshwater ecosystems (Boyce et al. 2010;Murphy et al. 2020;Borics et al. 2021). These researchers have concluded that the dynamic changes of zooplankton in terms of genera richness, diversity, and density are instigated by ecological characteristics such as biotic and abiotic factors and their interactions.
In Malawi, Lake Malombe, also described as a permanent floodplain lake, is located in the Upper Shire River Basin in the outlet of Lake Malawi in Mangochi District (FISH 2015). The Upper Shire River Basin lies entirely within the Great African Rift Valley system and is characterized by major and minor tectonic movements. It is the largest in Malawi and the fourth-largest in Africa. Lake Malombe is fed by water from Lake Malawi through an 18 km stretch of the Upper Shire River basin. The lake shares critical unique aquatic ecology with Lake Malawi, including rich fish biodiversity, endemism, and genetic plasticity (Kolding et al. 2019). Unlike Lake Malawi, Lake Malombe is a shallow, turbid, and nutrient-rich lake with shelving vegetated shores. It is designated as the most productive Lake in Africa (Tweddle et al. 1995) due to its enriched waters from inflowing streams originating from highly populated catchment areas and recycling nutrients in the sediments due to its shallow depth. Plankton forms an essential component of the aquatic food web in this lake. Lake Malombe has been extensively studied since the 1960s by biologists, hydrologists, and ecologists (Jambo and Hecht 2001;Hara 2006;Anseeuw et al. 2011;Jamu et al. 2011;Hara and Njaya 2016;Makwinja et al. 2021). However, a few studies about the temporal and spatial distribution of plankton in this lake are published (Dulanya et al. 2014). Information about the spatial distribution of zooplankton in response to ecological dynamics in this lake is still lacking. Therefore, the main objectives of this article are: (i) to assess the distribution and abundance of zooplankton and Chl-a in a different transect of the lake and (ii) to determine the influence of the physical-chemical variables and Chl-a on zooplankton density in the lake. Attempts are made to answer the following questions: (i) Is there any spatial variation in Chl-a and zooplankton taxa, predominately between the middle, inlet, outlet, eastern part, and western parts of the lake? (ii) Is there any seasonal variation in the Chl-a amount and zooplankton density in Lake Malombe? (iii) Which physical-chemical variables influence the spatial and seasonal variation of zooplankton density in the lake?

Study area
Lake Malombe is approximately 470 m altitude and is situated between 14 21 0 to 14 45 0 south and 35 10 0 to 35 20 0 east ( Figure 1). It is part of the Great Rift Valley system and is the third-largest Lake in Malawi, 30 km in length and 15 km in width, with a total area of 420 km 2 , an average depth of around 2-2.5 m, and a maximum depth of 6 m. It lies in a broken depression running northwest from Lake Chilwa to Lake Malawi, parallel to the Shire Rift Valley. Lake Malombe catchment experiences a warm tropical climate with seasonal air temperatures ranging from 27-30 C degrees Celsius and an average annual rainfall of 692 mm. In exceptional instances, air temperatures go as high as 40 degrees Celsius. The lowest air temperatures are experienced in May, June, and July, while the highest air temperatures are registered between October and November. Lake Malombe population within a 10 km radius is estimated at 64,000 households, of which 80% of the population are predominately fishers. The upland Lake Malombe catchment is dominated by maize cultivation, which covers 90% of arable land, while rice is commonly grown in the lowland areas of the lake. Lake Malombe is a vital ecosystem both local and international because of its high fish species diversity that is a source of food and income to the local population, and it is an important foraging ground for migrating birds. Sampling sites were selected, taking into account the accessibility to capture habitat and environmental variations across the lake. The selected sites included five transects (Inlet, Western, Eastern, Middle, and Outlet) of the Lake (Figure 1) with distinct characteristics. Each transect had twenty sampling points selected randomly.

Sample collection and laboratory analysis
Water samples were collected in the CWDS (May to July) and HDS (September to November) from the georeferenced sampling sites inside the Lake (Figure 1). Sampling was not conducted during the rainy season due to excessive precipitation. Water temperature ( C), pH, electrical conductivity (lS/cm), and Dissolved Oxygen (mg/L) were measured in -situ using WTW MultiLine P4 multi-parameter probe, model (Multi 3430). Soluble reactive phosphorous-SRP, nitrite, and sulfate (PO À3 4 , NO À 3 , SO 2À 4 ), were determined using UV/Visible spectrophotometer (model T90) following Molybdenum blue method in conjunction with Ultraviolet Spectrophotometry for nitrite and soluble reactive phosphorous-SRP, Turbidimetric method for SO 2À 4 : Cation (Mg 2þ , Na þ , Ca 2þ , K þ ) and anions (Cl -, HCO 3 -, CO 3 2-) were determined using an Electric auto titrator (model #775) following Microwave plasma atomic emission spectrophotometry technique (APHA 1998). For quality control, the charge balance of major ions for all water samples was done using the following equation (Hamzaoui-Azaza et al. 2011).

Sampling and determination of chlorophyll-a
Water samples were collected using five automatic ISCO samplers (Teledyne Isco, Lincoln, NE, USA) (one at each sampling transect). The average sampling depth of the inlet was 3.8 m, eastern part 2 m, western part 2 m, middle 3.8 m, and outlet 2.8 m. The sampling points were made on the shores of the lake except those collected from the inlet, outlet, and middle. A total of 200 samples (500 ml each) were collected from the sampling sections during the HDS and the CDWS, with 100 samples collected from each season. Water samples were preserved in a cooler box at À4 C and transported to the laboratory for analysis (Ekere et al. 2019). The water samples were filtered in the laboratory using a Whatman GF/C filter type, 25 mm in diameter, at a residual pressure of 0.7 bar. Any visible zooplankton was removed from the filter using forceps. The filter was then folded together with algae inside, blotted with absorbent paper to remove most of the water, and then placed in a properly labeled clean container. The filter was then placed under cold conditions (in the freezer at -20 C) for 24 hours. Before measurements, the extracts were mixed thoroughly and centrifuged for about 10 minutes at 500 Ã g (where g is gravitation acceleration (g ¼ 9.81 ms À2 ). A filter fluorometer with an excitation wavelength of 430 nm (10 nm bandwidth) and an emission wavelength of 680 nm (10 nm bandwidth) was used to determine Chl-a. The fluorometer was calibrated first using a commercial solution of pure chlorophyll manufactured by Sigma, UK. The concentration of the solution (in 90% acetone) was determined spectrophotometrically using an extinction coefficient of 87.67 Lg À1 (Jeffrey and Humphrey 1975) at 664 nm against a 90% acetone blank. The calibration was carried out with different Chl-a concentrations covering all the linear ranges for the relationship between chlorophyll concentration and instrument output. Also, the maximum acid ratio was determined by measuring the fluorescence of the standard before and after acidification. The sample extracts from the centrifuge tubes were then transferred to the fluorometer cuvette by carefully pipetting and then measured against a 90% acetone blank. 0.2 ml 1% v/v hydrochloric acid was added in the corvette, appropriately mixed, left for 2-5 minutes, and then measured again against a 90% acetone blank. Chl-a was determined according to the equation of Holm-Hansen et al. (1965).
Where: K ¼ calibration coefficient ¼mg Chl-a per ml 90% acetone per instrument fluorescence units, F m ¼ maximum acid ratio F 0 = F a of pure chlorophyll-a standard, F 0 ¼ sample fluorescence before acidification, F a is sample fluorescence after acidification and V e is extraction volume (ml), V f is a filtered volume (L).

Determination of zooplankton
A clear plastic box was lowered in the water column at a discrete depth (the inlet 3.8 m, eastern part 2 m, western part 2 m, middle 3.8 m, and outlet 2.8 m) and raised to trap the zooplankton inside the box. The water was allowed to exit a small mesh net (70-100 mm mesh size) attached to the lower wall of the box, and zooplankton was collected inside a sampling bottle at the end of the net. Three sample replicates were collected at each sampling station. The water samples were preserved with a 5% buffered formaldehyde solution, kept in a cooler box at low temperature and then taken to the lab for analysis. The samples were labeled correctly using waterproof permanent Blank India ink. The sampling period was between October and November (HDS) and between May and June (CDWS). Sampling activities were replicated three times per season. In the laboratory, the sample was drained of excess formaldehyde. For tiny zooplankton, such as rotifers, nauplii, which occur at high densities (>1000 per liter), 100 ml concentrated samples were decanted using a hand pipette and reduced to 5 ml for microscopy. The pipette was rinsed with distilled water into a culture dish to remove any adherent organisms. The sample volume's subject percentage at high magnification multiplied by total volume and divide by counted volume was done to obtain the total number of animals in the sample. The zooplankton identification was done at the lowest possible taxonomic level following standard taxonomic references (Shiel 1995;Fernando 2002) and observed morphometric features using a calibrated compound light microscope (Max II 1202.4000 model) fitted with an ocular micrometer. For large zooplankton such as Diaphanosoma spinulosum, which occurs at relatively low densities (1to 100 per liter), the whole sample was scanned at low magnification, counting all observed individuals. The zooplankton density in the lake was estimated by dividing the number of animals filtered with the trap of each species by the volume of water.

Data analysis
Statistical analysis was done using SPSS 20 and Microsoft Excel windows 2016. Before analysis, the physical-chemical variables (except pH), the Chl-a, and zooplankton density were subjected to a homoscedastic test to check whether the data were normally distributed or not (Arribas et al. 2014). If the data was not normally distributed, then log-transformation was done to achieve the homoscedasticity for inferential analysis (Pek et al. 2018). The differences of physical-chemical variables, phytoplankton biomass, and zooplankton density were analyzed in different seasons using paired t-test. The Canonical Correspondence Analysis (CCA) analyzed relationships between Chl-a, zooplankton density, and physical-chemical variables (Denis 2019). The CCA was chosen because of its ability to extract significant gradients among combinations of explanatory variables in a dataset. Pearson correlation analysis confirmed the significant relationship between physical-chemical variables, the Chl-a, and the zooplankton density.

Results and discussion
The majority of the Lake Malombe local population depends on unprotected water from the Lake (Kalumbi et al. 2020). As part of the lake catchment, water from rivers that drain into the lake is heavily contaminated with agricultural chemicals such as inorganic fertilizers, insecticides, and other human contaminants (Otu et al. 2011), although studies focusing on the limnological status of these rivers are scarcely available. Over 90% of the local population further depends on the fishery as the primary source of livelihood ). This study used nutrient concentration (nitrite, SRP, and sulfate), major cations and anions (Ca 2þ , Na þ , Mg 2þ <K þ, HCO 3 -, Cl -, CO 3 2-) , pH, WT, DO, EC, TDS, Turbidity (NTU), and Chl-a in Table 1 as variables to explain the ecological status that influences the spatial and seasonal distribution of zooplankton density in this lake. These major ions were selected in this study because their concentrations are currently exhibiting an increasing trend in the world's freshwater ecosystems (Simmons 2012). The pH plays a significant role in explaining chemical speciation, dissolution, and precipitation Nkwanda et al. 2021). The decrease in pH indicates increased metal dissolution in the lake and a high probability of bioavailability and toxicity to humans and aquatic biota (Magalhães et al. 2015;Bhateria and Jain 2016). The increase in carbon dioxide lowers the pH because carbon dioxide reacts with water to produce carbonic acid (Nazari et al. 2010). Again, the high rate of respiration and low photosynthetic activity results in low dissolved Oxygen and more CO 2 produced, which eventually lowers pH (Wang et al. 2017)-a scenario observed during the CDWS. The lake inlet recorded the highest pH during the CDWS due to the increased inflowing carbonate-rich water from Lake Malawi. Because of the hydrological connection between these two lakes, Lake Malombe behavior, including pH, is strongly influenced by Lake Malawi (Dulanya et al. 2014).
On the other hand, the highest pH recorded in the eastern section of the lake during the HDS (Figure 2) is attributed to increased photosynthetic activity instigated by increased aquatic plants in this section of the lake than the rest. The values of pH (8.94-9.19) measured in this study were in the side of alkaline with seasonal variations and were within recommended values for a healthy freshwater aquatic ecosystem (WHO 2011;EPA 2020). Lake Malombe is one of the climate-sensitive lakes due to its shallowness. Climate variability has a significant effect on the behavior of the lake. Water temperatures vary with seasons and in different parts of the Lake (Figure 2). The highest WT  (1992) SRP means soluble reactive phosphate, CDWS means cool dry windy season, HDS means hot dry season, WT means water temperature, MRL means maximum recommended limit. recorded at the inlet was attributed to the inflowing water from the Shire River into the lake. The water temperature, pH, and DO had negative mean paired differences and were significantly (p < 0.05) higher during the HDS than the CDWS (Table 2). Nitrite, sulfate, and phosphate occur naturally from the weathering process. The study showed a spatial and seasonal variation of nitrite and SRP concentration in the Lake (Table 1). The mean concentration of nitrite was lower at the western part of the Lake during HDS and CDWS (Figure 2). The concentration increased significantly (p < 0.05) during the CDWS as compared to HDS (Table 2). Van Loon and Duffy (2005) suggested that excessive NO 3 is highly linked to human health problems such as methemoglobinemia, gastric cancers, goiter, and hypertension. Nitrite was found to be higher than the Unted Nations and EPA recommended concentration (0.06 mg/L) for healthy freshwater ecosystems (United Nations 1992; EPA 2020) ( Table 1). Studies have shown that nitrite becomes less toxic in the presence of calcium and chloride. However, the concentration of these significant ions reported in this study is much lower (18.74 meq/L, 20.13 meq/L) than the recommended The Environmental Protection Agency report also indicates that nitrite above 0.06 mg/L is generally toxic to fish and can convert hemoglobin to methemoglobin causing pathological changes in fish organs and tissues under continuous exposure (EPA 2020). Steinman et al. (2014) also noted that fish species such as catfish and tilapias are susceptible to nitrite concentration. This suggests that increased nitrite concentration during CDWS (0.83 mg/L) and HDS (0.27 mg/L) reported in this study is worrisome to the aquatic biota. The turbidity was high during CDWS as compared to HDS. The increased turbidity during the CDWS was attributed to increased wind erosion, sand mining, and farming activities (Razali et al. 2018;Woldeab et al. 2019;Nkwanda et al. 2021). Studies conducted by Jamu et al. (2011) and Palamuleni et al. (2011) revealed that there had been a significant land use change since the 1980s in the west part of Lake Malombe catchment due to increased agricultural activities, and these changes are linked to many environmental alterations (Simmons et al. 2008) such as increased gullying of the catchment, culminating in significant sedimentation of the lake littoral plain.
The presence of sulfate in a very high concentration is also associated with respiratory problems, and in conjunction with sodium and magnesium, sulfate can exert a purgative effect on digestive tracts (Quattrini et al. 2016;Nkwanda et al. 2021). The paired t-test (Table 2) showed that the SO 4 2concentration during CDWS and HDS was not significantly (p < 0.05) different. The middle section of the lake recorded the highest SO 4 2concentration during the HDS, while the inlet recorded the highest during the CDWS (Figure 2). The increase in SO 4 2concentration during the HDS at the middle section of the lake is attributed to prolonged hydraulic water residence time that allows accumulation of industrial wastes and other salt deposits from an inflowing network of rivers into the lake (Bhateria and Jain 2016). Another reason was increased evaporation and groundwater inputs since the lake is also fed by groundwater during the hot, dry season when the water level decline and evaporation increases.
On the other hand, the inlet recorded the highest during the CDWS because of shorter hydraulic water residence time and increased inflow of water from a network of rivers within the catchment that carry salt deposits into the lake (Chandra et al. 2012). However, the SO 4 2value recorded in this study was below the global freshwater values and WHO guidelines for freshwater (WHO 2004). The low SO 4 2concentration in Lake Malombe was because African Great Lakes such as Lake Malawi are known for having low sulfate concentrations (Hecky and Bugenyi 1992). Most of the water in Lake Malombe comes from Lake Malawi through an 18 km stretch of the Upper Shire River. Some studies have suggested that the SO 4 2in African Great Lakes is low enough to limit phytoplankton growth, for example, Lake Victoria, though more recent experiments indicate this is not the case (Ndebele-Murisa et al. 2010). The classical model of lake eutrophication explains that the biogeochemical processing of phosphorous varies with the season (Robson 2014;Wentzky et al. 2018). This indicates increased anthropogenic activities, nutrient input from the surrounding watershed, and climatic forces that influence water residence time and the lake's productivity by increasing algal growth, bacterial metabolism, and nutrient recycling rates during the HDS.
The SRP was the highest at the outlet during CDWS and lowest during the HDS (Figure 2). The increased SRP at the outlet is attributed to increased water current during the CDWS (Li et al. 2013). Another factor is increased sediment phosphorous release (Li et al. 2013), which is more pronounced in the Lake outlet than the rest of the lake section. Inlet, on the other hand, recorded the highest SRP during the HDS as compared to CDWS. The increased SRP concentration in the inlet is attributed to increased weathering rocks, wastewater, fertilizers from cropland, and run-off (Bhateria and Jain 2016). The concentration of SRP further exponentially increased (Figure 3(a)) with an increase in water depth. The bottom of the lake also displayed a higher value of SRP, which was attributed to P release from the bottom sediments (Golterman 2004). Kowalczewska-Madura et al. (2018) suggested that phosphorus from the bottom sediments in shallow lakes such as Lake Malombe can contribute 99% of total Phosphorous in the lake. Ndebele-Murisa et al. (2010), on the other hand, suggested that primary production is significantly influenced by nutrient concentration, and these, in return, are affected by thermal stratification, which is common in many tropical African lakes such as Lake Malombe. Figure 3(b) shows that Chl-a in Lake Malombe increased rapidly with an increase in SRP at the beginning and reach the peak concentration at 2.25 lg/L, which suggests that Chl-a can be limited by SRP (Dzialowski et al. 2005). The results conform to the findings of Çelik (2013), who reported a significant correlation (r ¼ 0.77) between Chl-a and SRP in Çayg€ oren Reservoir, Turkey. Lee and Liu (2018) also noted that SRP is the biologically available nutrient for maintaining primary productivity, and its concentration increases along with the scent of the trophic states. Similar results were reported by Hudson et al. in 56 lakes in North America (Hudson et al. 2000). Table 1 further showed higher oxygen saturation during the HDS than during the CDWS. The higher DO saturation during the HDS was due to increased photosynthetic activities. The SRP level reported in this study was further higher than those reported in Lake Malawi, suggesting that Lake Malombe is a nutrient-rich lake compared to Lake Malawi, which is attributed to its shallowness and inflowing nutrient-rich water from its highly populated catchment. The Chl-a was high during the HDS as compared to CDWS. The high Chl-a is attributed to the availability of nutrients such as SRP during the HDS compared to CDWS. The western section of the lake recorded the highest Chl-a during the HDS, and the outlet recorded the highest during the CDWS (Figure 2). The variation is attributed to nutrient loading from the upstream rivers that drain into the lake (Marcarelli and Wurtsbaugh 2007). In 1991-1992, the "GOM/FAO/UNDP, 1993" project reported 4.6 mg/L Chl-a concentration in Lake Malombe. However, this concentration is slightly higher than the mean range (3.11 mg/L À3.92 mg/L) reported in this study during the CDWS and significantly lower than the concentration (11.13 mg/L À13.36 mg/L) reported during the HDS, suggesting that the lake becomes more eutrophic during the HDS than CDWS.
The charge balance of both cations and anions for all water samples calculated from the individual cation and anion concentrations (Table 1) was low (less than 10% on a morality basis). The order of major cations and anions is Ca 2þ <Na þ <Mg 2þ <K þ < HCO 3-<Cl -< CO 3 2-< SO 4 2respectively. The paired t-test in Table 2 shows that the concentration of major ions such as Na þ , K þ , Cl -, and CO 3 2significantly (p < 0.05) varied with season. The concentration of these ions except HCO 3 was higher during the HDS than CDWS. However, these concentrations of major ions (except Ca 2þ and Mg 2þ ) were higher than those reported in Lake Malawi (Bootsma and Hecky 1993;2003). The increased concentration of these ions in Lake Malombe was attributed to increased sediment loss from the lake catchment due to increased anthropogenic activities and weathering processes instigated by increased temperatures and wind speed. According to Abubeker (2017), sediment loss induced by human activities around Lake Malombe catchment is estimated at 10 tons per hectare annually. On the other hand, a significant decrease in calcium and magnesium concentration (below values reported in Lake Malawi by Hecky et al. 2006) was attributed to the increased snail population (Kamtambe et al. 2019), high evaporation rate instigated by rising atmospheric temperatures (El Gammal et al. 2017) and longer hydraulic residence time which allows the possible accumulation of ions in the sediments, Hecky 2003, Jeong et al. 2007;Saleem et al. 2015). Table 2 further showed that the concentration of the major elements recorded in Lake Malombe was below the global freshwater values (Berner and Berner 1987).

Spatial variation of zooplankton density and its interaction with physical-chemical variables and chl-a
Zooplankton species undergo physiological changes to survive different aquatic systems (Richardson 2008). The season changes affect species number, diversity, biomass, and diversity as well as community structure. The seasonal change of phytoplankton biomass within a year also influences zooplankton bottom-up effects and top-down effects through the aquatic food web (Doi et al. 2013). With the current study site being an inland closed-basin lake with some sampling sites located further away from the discharge point of the main Shire River (inshore areas), some differences in zooplankton density were predicted. The zooplankton density exhibited strong spatial and seasonal variations (Figure 4). Similar to other lakes in temperate and tropical regions, zooplankton density was dominated by copepods (T. cunningtoni, Nauplii, M. aeq. aequatorialis), Cyclopidae (T. neglectus), Cladocera (B. longirostris), Ctenopod (Diaphanosoma excisum), and Rotifers. T. cunningtoni is also found in Lake Malawi (Ngochera and Bootsma 2011) and its presence in Lake Malombe is linked to the fact that these two lakes share some unique characteristics of the larger lake's aquatic ecology. The Eastern part of the lake had the highest population of dominant zooplankton species (T. cunningtoni, M. aeq. aequatorialis, T. neglectus, B. longirostris, Diaphanosoma spp, Nauplii, and Rotifers) during the CDWS (Figure 4) which was attributed to nutrient influence and water Physico-chemical characteristics. For example, the eastern part of the lake is relatively shallow with highwater temperatures, pH, and lowest nitrite, creating a conducive environment for the growth and distribution of zooplankton species. The increase in zooplankton species in the eastern part of the lake is further evidenced by a decrease in Chl-a, suggesting that the large population of zooplankton species highly grazes the phytoplankton population. The HDS also recorded high zooplankton species in the eastern part of the Lake (T. cunningtoni, M. aeq. aequatorialis, T. neglectus, B. longirostris, Diaphanosoma spp., and Nauplii), which is attributed to the availability of food. On the other hand, the high Rotifer population reported at the lake outlet during HDS is attributed to increased water current and high dissolved Oxygen. The increase in zooplankton density during the HDS can further be explained by increased Chl-a amount from HDS to CDWS, and this mainly occurs in eutrophic lakes such as Lake Malombe, where climate change lengthens cyanobacterial growth season (Schagerl and Oduor 2008;Paerl and Paul 2012). Nauplii require a much higher food concentration for development than copepods (Saiz et al. 2014). Their low ingestion rates indicate that low feeding efficiency can be the reason for their high food demand (Sinistro 2010). Under conditions of food scarcity, for example, during the CDWS, where Chl-a is significantly lower, as seen in Table 2, some Nauplii spp have a lower survival rate (Mathews et al. 2018). This explains the reason for their low abundance during the CDWS. A similar observation is made in Rotifer. The positive correlation coefficient (Table 3) explains that the lake's Chl-a amount stimulated the increased zooplankton population. Hallegraeff (2010) suggested that phytoplankton is the primary producer in the complex food web system and this phytoplankton also produces Oxygen and organic matter through photosynthesis. This mechanism provides a better connection between primary producers with more advanced consumers in the food webs and provides a crucial link in the aquatic food web system (Li et al. 2020). The changes in zooplankton density as seasonal changes are predominately influenced by physicalchemical (bottom-up effects) and predation (top-down effects) through the aquatic food web (Lv et al. 2011;Muhid et al. 2013;Li et al. 2020). Figure 5 shows the canonical correspondence analysis (CCA) of zooplankton density and water physical-chemical variables and Chl-a. Pomati et al. (2020) noted that the Chla amount is strongly linked to the total nutrient concentration (SRP and nitrite) available in the system. In this study, the arrow length indicates the significance of the variable and shows positive or negative correlations with the axis. The percentage of variance and Eigenvalues of each site on axis 1 was higher than axis 2. At Lake Malombe West, CCA was drawn between 14 physical-chemical variables and Chl-a and the dominant zooplankton density. The eigenvalue for axis 1 (0.009) explained 98.3% correlation and axis 2 (0.002) explained 22.36% correlation between physical-chemical variables, Chl-a, and dominant zooplankton. The pH and DO showed a close relationship with the zooplankton community, which shows high productivity in the western part of the lake. Nauplii, R, and BL positively correlated with axis 1 and indicated a positive correlation with Mg 2þ, pH, SO 4 2-, HCO 3-, Nitrite, SRP, Ca 2þ, and Chl-a. However, DS, TC, TN, MAA were positively correlated with axis 2 and further correlated with CO 3 2-WT, TDS, DO, Cland Na þ . The population density of DS, TC, TN, MAA were negatively affected by Mg 2þ , pH, SO 4 2-, HCO 3-, Nitrite, Ca 2þ , Naupilli, R and BL were negatively affected by CO 3 2-, WT, TDS, Cl -. At the Lake Malombe outlet, 14 physical-chemical variables and Chl-a and the dominant zooplankton density were considered to draw CCA. Lake Malombe outlet had the most abundant zooplankton species. On axis 1, Eigenvalue (0.022) explained 80.37%, and Eigenvalue of axis 2 (0.005) explained 19.63% correlation between physical-chemical variables and Chl-a and dominant zooplankton density. The R, TN, TC, DS showed a positive correlation with xis1, which indicates the effect of Cl -, CO 3 2-, HCO 3 -, SO 4 2-, DO, SRP and Nitrite. Naupilli, MAA, BL showed a positive relationship with Axis 2 and positively affected by WT, pH, Ca 2þ , Mg 2þ , Na þ , K þ , and CO 3 2-. At Lake Malombe middle, the Eigenvalue at axis 1 was (0.0169) and explained 93.31% of the correlation while Eigenvalue at axis 2 was (0.001) explained 6.69% correlation between physical-chemical variables and Chl-a and dominant zooplankton density. Lake Malombe inlet, on the other hand, had an Eigenvalue of axis 2 (0.001) and explained 76.4% of the correlation. The R, MAA, and DS correlated positively with axis 1 and HCO 3 -, DO, SO 4 2-, Cl -, Chl-a, and CO 3 2were positive while Nauplii, BL, TC, and TN were correlated positively with axis 2 and with pH, nitrite, and TDS. Figure 5 further shows that Lake Malombe east had Eigenvalue (0.009) and explained 88.1% correlation at axis 1. On the other hand, axis 2 had Eigenvalue (0.0013) and explained 11.9 of correlation. TN, DS, R positively correlated with axis 1 while BL, Nauplii, TC, MAA positively correlated with Figure 5. CCA biplot (at Lake Malombe West, outlet, middle, inlet and east) between physic-chemical variables and Chl-a and zooplankton density. axis 2. The Ca 2þ , Chl-a, CO 3 2-, Mg 2þ , Na þ , Cland DO positively correlated with DS, R, TN while BL, Nauplii, TC, MAA correlated positively with Nitrite, SRP, SO 4 2-, pH, and HCO 3 -. Na þ also showed a positive relationship with TC, MAA, DS, Nauplii, and Rotifers.
Pearson correlation analysis results in Table 3 confirmed that zooplankton species density in the lake followed distinct sequences regulated by the water physical-chemical variables such as WT, pH, DO, nutrients. Karl Pearson's correlation between WT, SRP, nitrite, pH, DO, Chl-a, and zooplankton density in Table 3 were positive. El Gammal et al. (2017) acknowledged that the pH range within (8-9) is appropriate for the development of zooplankton. The Clalso displayed a positive relationship with D. excisum and Rotifers and had a correlation coefficient of (r ¼ 0.488; r ¼ 0.472). The CO 3 2and TDS showed a negative relationship with B. longirostris, Nauplii, Rotifers (r=-0.423; r=-0.535; r=-0.506). The factors depicted in this study may not be the only factors influencing the zooplankton population variation in Lake Malombe during the CDWS and HDS. Predator abundance and the absence of predators also play an important role. For example, there is general evidence that crustaceans can effectively influence the Rotifer population via predation and competition. In the case of Lake Malombe, the dominant crustacean is B. longirostris. Rotifer could also be the main food source for Cladocera, cyclopoid, copepods, calanoid, and predatory rotifers, though this kind of predator/prey relationship is ephemeral as rotifers evolve many defense mechanisms against invertebrate predators. The upper trophic levels could also influence nutrients availability for primary consumers via the stoichiometry of nutrient recycling. Studies have shown that C: N and C:P ratios vary among the zooplankton species (Rhee and Gotham 1980;Van Nieuwerburgh et al. 2004). For example, Cladocera has a higher proportion of phosphorous and a lower proportion of nitrogen than copepods (Sommer and Sommer 2006). This differential nutrient uptake modifies the proportion of nutrients in the water column during both HDS and CDWS. Karl Pearson's correlation in Table 3 demonstrates a positive significant correlation coefficient between WT and Diaphanosoma spp, T. neglectus, Rotifers, and T. cunningtoni, suggesting that rising WT reported during the HDS is associated with the increase in the population of these species. Similar findings have been reported in other African tropical lakes such as Lake Chad and Lake George, Lake Kariba (Talling 1986;Moyo 1991). As observed in other studies, the low temperature can lead to a decrease in zooplankton density in the Lake (Brucet et al. 2010;Cremona et al. 2020;Li et al. 2020;Pomati et al. 2020). In early HDS, Chl-a develops when nutrients are available and when temperatures are high. Under warmer water temperatures, Chl-a increases. Some zooplankton groups become sensitive to water mixing; therefore, their dominance in HDS can be explained by the thermal stratification phenomena which prevent water mixing.

Conclusion
In this study, the Chl-a and zooplankton density was high during the HDS compared to CDWS. The highest biomass of Chl-a amount recorded during the HDS was mainly attributed to the relatively high WT, pH, and SRP concentration. The study identified that copepods (T. cunningtoni, Nauplii, M. aeq. aequatorialis, Cyclopidae (T. neglectus), Cladocera (B. longirostris), Ctenopod (D. excisum), and Rotifers as dominant zooplankton. T. cunningtoni are more dominant in HDS when nutrients are available, and WT is high. Interestingly, the results further revealed that WT, pH, DO, nutrient concentration, and some significant ions, as well as Chl-a, were the most important variables influencing the zooplankton density in the lake. The findings in this study are essential for the management of Lake Malombe and other tropical inland freshwater shallow lakes with similar attributes across the globe. The knowledge generated from this study is critical in fishery management and understanding the ecological balance of tropical inland lakes.
Coordinator of the African Union Aquaculture Working Group and Coordinator of the NEPAD Fish Node of the Southern Africa Network on Biosciences (SANBio).
Dr. Tena Alamirew has over 25 years of professional experience in academics, research, and leadership. He has lectured numerous courses in the areas of water & soil for undergraduate & graduate students and supervised six Ph.D. & over 60 MSc students. In addition, he has been involved in the design and effective implementation of several national and international collaborative research projects, much of which is used to support graduate students; he demonstrated his leadership attributes while was serving as the academic and research Vice of Haramaya University for about seven years, and as founding Director of Ethiopian Institute of Water Resources. Dr. Tena is now serving as the Deputy Director of Water and Land Resource Centre and heads the center's research division.

Data availability statement
The datasets analyzed during the current study are available from the corresponding author on reasonable request.