Water constitutes part of the dynamic aquatic life-supporting system in which organic and inorganic constituents are dissolved or suspended and in which a wide  variety  of  organisms  live  and  interact  with  each other (Awah, 2008). In addition, water bodies provide valuable ecosystem services, such as water supply, production, recreation and aesthetics. Having it available in   sufficient   quantity   and   quality  contributes   to   the maintenance of health. Meanwhile, anthropogenic activities deteriorate surface waters (Chukwu et al., 2008) and impair their basic use through the process of pollution (Wu, 2005; Hur and Jung, 2009; Zhang et al., 2009).

Study site

Buea is situated between latitude 3°57^’ and 4°27^’N,  longitude  8°58^’and 9°25^’E, and at an elevation of about 500 to 1000 m above sea level on the east flank of Mount Cameroon. The mean annual precipitation and temperature stand at about 3000 mm and 28°C, respectively. The relative humidity is 86% and sunshine ranges from 900 to 1200 h per annum (Folifac et al., 2009). The climate is equatorial, with two seasons: A dry season from November to February and a rainy season from March to October.  Buea is a watershed area, characterized by a number of springs some of which develop into streams at lower elevation. Less than half of the population of this municipality has access to pipe-borne water, which is also erratic particularly in the dry season, increasing pressure on open and risky sources.  There is therefore chronic portable water scarcity due to poor management of the available sources, placing the inhabitants at risk (Agbor and Tefeh, 2013) of water related diseases.

Sample collection and analysis

Nineteen points on different springs and streams were sampled (Figure 1 and Table 1) in March (transition between the dry and rainy season) 2013, based on anthropogenic interactions:  Points exposed to car wash (CW, 3), those subjected to municipal wastes deposition (MW, 4), that subjected to both car wash and municipal wastes deposition (CM, 1) and those exploited for household usage including drinking (DW, 11). The number of sampling sites per category was determined by availability of such sites within the study area. Farming as an activity cut across all the categories.

All drinking water sources were springs, seven of which had been constructed and water channelled through pipes while the others were open sources. Of these springs, one (Man Ndongo) developed into a stream that runs across a  major  residential  area,thus considered DW at the source and MW downstream. The other streams in this study had more remote sources, flow through residential areas and are subjected to different anthropogenic influence. All 19 sources were very shallow with no possibility of stratification (Plate 1).  At each sampling site, two sets of water samples (one for chemical analysis and the other for phytoplankton analysis) were collected in triplicates, in 50 mL sterilized plastic bottles following standard procedures (Bellinger and Siegee, 2010). The six DW samples were either collected directly from the pipes or 5 cm below the water surface (in the case of the open sources) at six different points. The CW and CM were sampled at the car wash points because these are point pollution sources while the MW, being non-point pollution sources were sampled along the stream. The CW and CM samples were collected at six equally spaced positions across the stream breadth while MW samples were collected 50 m apart along the stream course. For the six samples per site, three were bulked and subsampled for algal analysis while the other three were similarly bulked and subsampled for nutrient analysis.

Before collection, in-situ measurements were recorded at these points: pH and temperature were measured using a portable Hanna H198127 pH/temperature meter (pH/°C), electrical conductivity was measured using a conductimeter (Hanna H198303) in µS/cm and total dissolved solids were measured in mg/L using a TDS tester. At each pipe, a 10 cm plastic bowl was filled, the probe was inserted 5 cm deep and the insitu-reading noted.

Each sample for phytoplankton analysis was treated with three drops 10% Lugol’s Iodine and transported in  ice  containers  to  the Life Sciences Laboratory of the University of Buea for phytoplankton analysis. Nitrate, ammonium, sulphate and phosphates in water samples were analysed at the Soil and Environmental Chemistry Laboratory, University of Dschang, Cameroon.

Phytoplankton identification

Slides of each sample were prepared in triplicate. A drop of sample was placed on a sterilized slide covered with a slip and observed under the microscope.  Phytoplankton species were identified and counted by  the use  of  a  binocular  light  microscope  (Olympus  BH2),  at  a magnification  of 1000 x. Algae were scored for absence (-) and presence (+) in the different water sources for each category. The number of particular alga in the mount was also noted. Identification was through comparative morphology and description using relevant text books, manuals and articles (Trégouboff and Maurice, 1957; Compère, 1977; Nguetsop et al., 2007; Bellinger and Siegee, 2010). Algae were classified according to algaebase.org.

Statistical analysis

The abundance of each alga per milliliter was obtained from the sum  of  its  occurrences  in  the  three  slides  (drops)   as   follows:

Phytoplankton occurrence and abundance

Overall, 66 phytoplankton species belonging to 44 genera, 34 families and six phyla were identified from the nineteen sampled points (Table 2.). Of these, nine species were cosmopolitan while 22 (Table 3), three (Achnanthes sp, Cocconeis sp and Stauroneis product) and two species (Cosmarium moniiforme and  Prorocentrum minimum) were unique to CW, MW and DW categories respectively (Figure 2). There was no  species  unique  to CM category. Nitzschia was the most abundant genus while the Bacillariaceae was the most abundant family.

Bacillariophyta was the most abundant phylum. The occurrence of these algae was however not uniform among categories and sites, with mean number of phytoplankton species per category being 17, 8, 11 and 4 respectively for CW, MW, CM and DW (Table 2). Nitzschia sp. (22.48 %), was the most abundant in CW category while, Chlorogonium sp., Elakatothrix sp., Nitzschia seriata and Anacystis sp. were the least abundant with 0.29% each (Table 3). In the MW category, Chlorella sp. scored 27.57% while Achnanthes sp., Anabaena sp., Cocconeis sp., Microcystis sp., Navicula cucpidata, Pinnularia sp. and Stauroneis producta were the least abundant with 0.54% each (Table 4). The CM category  had  Nitzschia  as  the  most abundant (34.48%) while Ceratium, Cosmarium, Oscillatoria and Tetraëdron were the least abundant with 1.72 % each (Table 5). Chlorella was the most abundant (24.27 %) in the drinking water category, while Amphipleura pellucida, Closterium abruptum, Hantzchia amphioxys, Nitzschia constricta, Pediastrum duplex, Prorocentrum minimum and Synedra ulna were the least (0.22% each) (Table 6).

Diversity and similarity of phytoplankton within and between categories

The different  water  sources  had  very  little  similarity  in phytoplankton composition with very few species common to all sites within a category. In the drinking water sources for example there was no species that occurred in all the sites. Algal diversity was generally high within water sources in each category with Simpson’s diversity indices ranging from 0.704 in CWEK (Car wash) to 0.896 in MWDN (Municipal waste) (Table 7). There were however no significant differences in diversity between the categories of water sources.

Water physico-chemical parameters and their relationship with phytoplankton diversity and abundance

There  were  variations  in  physico-chemical  parameters among sites and categories. In the car wash category, the highest phosphate (0.73 mg/l), ammonium (3.36 mg/l) and nitrate (1.92 mg/l) were from CWNA. CW18 had the least phosphate (0.61 mg/l) while CWEK had the least ammonium and nitrate (0.56 and 0.89 mg/l, respectively). Temperature and pH were highest in CWEK (23.1 and 7.8 respectively) and lowest in CWNA (22.1) and CW18 (7.3) respectively. The electrical conductivity and total dissolved solids were highest in CW18 (255 µS/cm and 125 mg/l respectively) and lowest in CWEK (205.7 µS/cm and 105 mg/l respectively).  In the municipal waste category, phosphates were highest in MWBB (0.80 mg/l) and lowest in MWMP (0.59 mg/l). Ammonium was 0.67 mg/l for all sites except MWTB with the least value of 0.53 mg/l. Nitrates were highest in MWTB (1.32 mg/l) and least in MWMP (0.87 mg/l). Sulphates ranged between 0.16 mg/l (MWBB and MWMP) and 0.13 mg/l (MWTB). Temperature was in the range 21.4 (MWBB) and 24.1 (MWDN). pH ranged between 6.6 (MWBB and MWTB) and 7.7 (MWDN)  Electrical conductivity and total dissolved solids were highest in MWBB (226.7 µS/cm and 108.3 mg/l, respectively) and least in MWTB (141 µS/cm and 73.3 mg/l, respectively). In the drinking water category, phosphates were highest in DWWO and DWWN (0.75 mg/l) and lowest in DWBB (0.45 mg/l). Ammonium was highest in DWWO (0.83 mg/l) and lowest in DWSP (0.57 mg/l). Nitrates were highest in DWNJ (1.43 mg/l) and lowest in DW18 (0.62 mg/l). Sulphates were highest in DWWO (0.15 mg/l) and lowest in DW16 (0.11 mg/l). Temperature ranged between 20.5 (DWSP) and 22.9 (DWBB). pH was highest in DWWN (7.5) and lowest in DWSP (6.1). Total dissolved solids ranged between 74.7 mg/l (DW18) and 141.7 mg/l (DWBS) while electrical conductivity was in the range 146.3 µS/cm (DWWN) and 288.3 µS/cm (DW18). Mean values of the different parameters for the different categories ranged between 0.65 mg/l (DW) to 0.71 mg/l (MW and CM) for phosphates, 0.64 mg/l (MW) to 2.24 mg/l (CM) for ammonium, 1.08 mg/l (DW) to 1.53 mg/l (CW) for nitrates, 0.13 mg/l (DW) to 0.14 mg/l (other categories) for sulphate, 21.8 (DW) to  22.9 (CM) for temperature, 6.9 (other categories) to 7.6 (CW)  for pH, 197.4 µS/cm (MW) to 233.3 µS/cm (CM) for electrical conductivity and 92.3 mg/l (MW) to 111.8 mg/l (DW) for total dissolved solids.

In this study the phytoplankton occurrence in the water sources was in the order Bacillariophyta > Charophyta > Chlorophyta > Cyanobacteria > Miozo > Euglenophyta. These findings are similar to those of Sorayya et al. (2011)  and  Wladyslawa   et   al.  (2007),   who   reported Chlorophyta, Bacillariophyta, Cyanobacteria and Dinophyta as dominant in the fresh water communities. Similarly, Laskar and Gupta (2009) recorded 34 phytoplankton taxa belonging to Chlorophyceae, Cyanophyceae, Bacillariophyceae and Euglenophyceae, in Chatla floodplain lake (India).

Waste discharge into water and fertilizer applications around water sources increase nitrogen and phosphorus levels in systems (Fonge et al., 2012). Also untreated car wash effluents have been reported to contain phosphates and nitrates above limit (Aisling et al., 2011) and these are nutrients which encourage algal growth. In this study, nutrients correlated positively with phytoplankton abundance while diversity correlated negatively with nutrients in agreement with the fact that increase in nutrients reduce diversity but increase abundance of tolerant species (Chislock et al., 2013; Fonge et al., 2015). Nitrates ranged from 1.08 to1.53 mg/l and were all below the 45 mg/l limit (WHO, 2008) while the phosphates ranged from 0.65 to 0.71 mg/L  and  were  all   above   the   0.30 mg/l   limit (WHO, 2004). The phosphates were more likely implicated in algal abundance. The higher levels of nitrates and phosphates in car wash activity explain the association between phytoplankton abundance and car wash category. The high number of unique species for the car wash category suggests that effluents from this activity promote the growth of particular algae while the lack of unique species in CM17 is probably the consequence of strong interaction between car wash and municipal waste discharge, without ignoring the consequences of single site sampling for the CM category. The dominance of Chlorella (Chlorophyta), Nitzschia and Navicula (Bacillariophyta) in the different sites is attributed to their ability to adapt to a wide range of physicochemical parameters, and anthropogenic influences (Celekli and Kulkoyluoglu, 2006).

In this study the diversity of phytoplankton in all the sites was high, with little differences in mean Simpson’s diversity between categories. According to Wan (2010), healthy environments are typified by greater diversity of organisms than degraded ones. However, the observed  variability in diversity indices within categories could be explained by categorization based on dominant influence. The age of the catchments (Anderson et al., 2002) and slope through which the water flows are also important factors accounting for such differences in diversity. In the drinking water category for example, the lowest diversity was recorded in Sasse water (DWSA), a colonial catchment that has known little maintenance for close to a century, while Koke water (DWKO), a recently (< 20 years) constructed catchment recorded the highest diversity.

Activity interaction had a tremendous impact on algal diversity, with carwash + municipal waste categories having the lowest number of species (11) and no unique species, as oppose to car wash category with 52 species. It seems evident from this data that the washing of cars directly in streams is dangerous to the aquatic ecosystem, as it resulted in the development of 22 unique species, compared to only three and two respectively for MW and DW categories and 9 cosmopolitan species for all categories. This probably   explains   the   low   overall    Sorensen similarity index of 0.40. The significant correlation between pH and abundance is justification of the impact of car wash on phytoplankton occurrence. However, the non-significant correlation between pH and diversity and between EC, TDS and diversity possible account for the uniqueness of the species in the different categories.The results confirm the complex relationship between diversity and environmental quality as proposed by Maznah and Mansor (1999). Chlorella, Chlamydomonas, Oscillatoria, Euglena, Navicula and Nitzschia were found repeatedly in all the categories and most sites. These genera are amongst those that have been reported in organically polluted waters (Jafari et al., 2006; Jafari and Gunale, 2006; Kshirsagar and Gunale, 2011; Kshirsagar et al., 2012 and Ayodhya, 2013). The presence of these algae in water bodies indicates eutrophic conditions (Fonge et al., 2012). In this study the Bacillariophyta dominated the pollution tolerant group, similar to observations by Arimoro et al. (2008), following studies in the Orogodo river in Nigeria. Their dominance in aquatic environments is a major indicator of water quality and environmental conditions because they are adapted to a wide range of physico-chemical conditions (Ajuonu et al., 2011; Fonge et al., 2012).