Distribution and Diversity of Macrophytes in Relation to Some Physico-Chemical Factors in the Ketar River, Ziway Catchment, Ethiopia

Distribution and diversity of macrophytes in relation to some physico-chemical factors in the Ketar River were studied from December, 2017 to November, 2018. Physico-chemical parameters and macrophytes were collected from three stations along the river for eight months. Onsite measurements and laboratory work of physico-chemical was analyzed as recommended by APHA [31]. Macrophytes were collected manually using belt transect method. Except for pH and surface water temperatures, all the physicochemical parameters measured showed no signicant difference spatially. During the study period, sixteen macrophyte species belonged to fourteen families were identied. Among the identied macrophyte, 11 of them were emergent, while 3 were rooted with oating leaves and 2 free-oating. Freeoating macrophytes were shared the highest abundance followed by emergent. This research observed that the site (site 3) that was exposed to minimal human impact was rich in diversity and abundance of macrophytes. All the sites were dominated by emergent macrophytes that attained the highest relative frequency followed by rooted emergent species. Azolla nilotica and Pistia stratiotes were shared the highest abundance and were the dominant macrophyte with the relative frequency of 7.24% and density of 40.91%, and 7.93% and 26.54%, respectively. Under a favorable environment, nutrient loading from nearby creates more favorable conditions for the infestation of the invasive species (A. nilotica and P. stratiotes) to ourish and out-compete the other species of macrophyte. Therefore, anthropogenic activities that enhance nutrient addition to the River should be regulated.


Introduction
The function of macrophytes in the ecosystems is related to their structural attributes like species composition, distribution, abundance, and diversity which in turn relies on various environmental factors such as light, water temperature, substrate composition, disturbance, competitive interactions, herbivory, epiphyte loading, water levels and water quality [1,2,3]. Sediment characteristics [4] and physical factors, such as slope, wind, or wave actions [1] are also important in determining aquatic vegetation patterns. In addition, competition between and within species have effect on the diversity and distribution of macrophytic species [5,6]. The distribution of macrophytes in freshwater ecosystems also seems to be in uenced by the nature of the geological formations and the degree and nature of the pressure exerted on these environments [7].
Aquatic macrophytes play an important role in the structure and function of aquatic ecosystems.
However, many threats to freshwaters (e.g., climate change, eutrophication, alien species introductions) will also result in reduced macrophyte diversity and favor the establishment of exotic species, at the expense of native species [8]. In Ethiopia, the invasive water hyacinth (Eichhornia crassipes) is distributed in different water bodies of the country [9] and it has created serious problems for the use of the water as a resource and may affect the abundance and diversity of other macrophytes [10]. Similarly, water lettuce (Pistia stratiotes) is among the world's worst weeds [11] and it decreased the growth of aquatic macrophytes [12]. Therefore, because of the signi cant role played by macrophytes in freshwater ecosystems and the introduction and spread of numerous nonnative species, understanding and quantifying their abundance, diversity and their relation to environmental factors is crucial for integrated management practices [13].
Macrophytes are an important component of many freshwater ecosystems that play different roles [13].
These include being primary producers [14], providing refuge for macro-invertebrates [15,16], zooplankton [17], and habitat for the feeding, breeding, and refuge of littoral sh [16]. Moreover, macrophytes affect the cycling of nutrients and contaminants [14,18], reducing shoreline erosion and sediment re-suspension. They can also be used as indicators of water quality [19,6].
In Ethiopia, the ecological importance of macrophytes has been neglected and only a few studies have been conducted in some Lakes in this regard. These include Unbushe [20] on the ecology of the wetland vegetation around Abaya and Chamo in Southern and Fincha'a-Chomen and Dabus in Western Ethiopia, Tamire and Mengistou [21] on macrophytes of Lake Ziway, Pattnaik [22] and Lalisa Gemechu [23] on macrophytes composition of Lake Hawassa. Kassa et al. [24] have reported on wetlands of Lake Tana and their macrophyte composition and Dida [25] on the oristic composition of wetland in Wonchi District, South Western Shewa. Recently, Wosnie et al., [26] and Getnet et al., [27] reported on the macrophytes of Lake Koka and the Gilgel Abay catchment, respectively. Rivers have been less investigated than lakes with regard to macrophytes. The aquatic macrophyte in Ethiopian Rift Valley lakes was documented by Kassaye et al., 2016 [28]. The Ketar River is one of the two important in uents which accounts for 62.7% of the in ow into Lake Ziway [29]. Along its course, the river carries a lot of nutrients and sediments downstream and the role of macrophytes in regulating their dynamics has not been explained. Hence, this study was conducted to understand the species abundance, distribution, and diversity, and to identify key environmental factors that drive macrophyte in the Ketar River which will provide vital information that can be used for management purposes.

Description of the study area and sampling sites
Ketar-Ziway watershed was named after Ketar River and Lake Ziway. Ketar River originates from the ridges of Kaka, Galama and Chilalo mountains along the South-eastern side of the watershed and ows in the western direction and forms part of Lake Ziway. The watershed is located within the Rift Valley basin between 7.3 0 and 8.2 0 North Latitude and 38.9 0 and 39.4 0 ' East Longitude. The Ketar catchment shows variations with altitude ranging from around 1638 m a.s.l. near Lake Ziway (at the inlet, present study) to about 4171 m a.s.l, on the high volcanic ridges along the eastern watershed (Chilalo and Galama Mountains) [30].
The river was studied at three sampling sites. The rst two sites are located at the upstream of the river and exposed to different human activities carried out in the watershed. The last one is located at the downstream of the river and is relatively less exposed to different stressors. The physical features of the sampling sites are summarized in Table 1. This site is located between agricultural land and transport system (regarded as environmentally better than site 1). There is a high probability of organic pollution and in ow of other excess agricultural inputs from upstream into this site as well. However, along the course of the river, there is relatively good coverage of macrophytes and is less impaired by human activities.

Sampling and Laboratory analyses
Water samples were collected from the surface of the river using polyethylene bottles for chemical analysis. Water samples were transported in an ice-box to the limnology laboratory of Addis Ababa University and analyzed immediately. The samples were analyzed following the standard methods described in APHA [31]. SRP and TP (after digestion with persulfate) were measured by the ascorbic acid method. Nitrate (NO 3 -N) was measured with the sodium salicylate method, while ammonia (NH 3 + NH 4 + -N) was determined by the phenate method. Nitrite (NO 2 -N) was determined by diazotization with sulphanilamide and coupling to Naphthylethylene diamine di-HCl. The concentration of total suspended solids (TSS) was determined gravimetrically after ltration of a known volume of water sample.

Sampling design of macrophytes
Samples along the River were sampled eight times (From October to May). Macrophytes were collected manually from three sites. These sites were selected based on their distance from human settlements and anthropogenic effect and presence of macrophyte coverage and accessibility for quantitative study [21]. Sites 1 and 2 are close to human settlements. Site 1 is more exposed human impacts. Even though site 2 is close to human settlements, the site is less impaired by human activities than site 1. Site 3 is far from human settlements and is minimally disturbed. It is well covered with different macrophytes that could be creating a sort of buffering system for the river. The aim of such site selection was to encompass varying environmental conditions based on their exposure to anthropogenic activities in the assessment of distribution and abundance of macrophytes and to note the variation in macrophytes distribution and abundance along the course of the River.
After collection, the macrophyte samples were rinsed in situ, blotted, pressed and transported to the National Herbarium, Addis Ababa University, Ethiopia, for identi cation. Identi cation was made to the species level using Ethiopian oras such as mentioned in the study by Hedberg and Edward [33] and Edwards et al. [34], herbarium collections at Addis Ababa University and with the help of standard kinds of literatures [35,36,37] and nally authenticated by the staff of the Herbarium The quantitative study was carried out in all study sites down the course of the Ketar River. To analyze the macrophyte community, a belt transect method was employed as recommended by IEP [38] and Dawson [39]. Each transect was taken from the shore bank perpendicularly towards the center of the River following Burlakoti and Karmacharya [40] and Dawson [39]. The number of transects at each site varied depending on vegetation cover and environmental heterogeneity [41]. The size of the quadrat used was 1 m 2 in all study sites, following the suggestion of Sutherland [42]. A total of 51 quadrates were laid during the study periods. The quadrates were laid along the transects at 50 m intervals following Dawson [39]. Macrophytes in each quadrat were counted by handpicking, and an independent morphological unit arising from rhizome was considered as an individual macrophyte as stated in the study by Pompeo and Moschini-Carlos [43].
The relative frequency and relative density of each species were calculated as in the study by Singh et al. [44] as follows: Relative frequency = (frequency of species A / total frequency of all species) X 100 Relative density = (density of species A / total density of all species) X 100 Relative abundance = (abundance of species A / total abundance of all species) X 100

Data analysis
The macrophyte data were quantitatively analyzed for abundance, relative frequency and relative density as in the study by Singh et al. [44]. Then the data generated were statistically analyzed using SPSS version 21. Spatial variations in abundance and diversity of macrophytes and physicochemical parameters were tested using one-way ANOVA followed by Tukey-HSD. After determining the gradient length (<2) using DCA (Detrended Correspondence Analysis), RDA (Redundant Analysis) was employed to determine the relationship between macrophyte species composition and abundance and environmental parameters using CANOCO for windows 4.5 version Software [45]. To reduce the effect of a rarity on RDA analysis, families that comprised <1% of the organisms at sampling sites were excluded [46]. The macrophyte species diversity index in the River was computed using PAST software.

Spatial variations in physico-chemical parameters
Water samples collected along the Ketar River for one year were tested for various physico-chemical water quality parameters ( Table 2). The measured pH values showed signi cant differences among the study sites (p<0.05), with the minimum and maximum mean pH values of 7.84 and 8.3 at sites 1 and 6, respectively ( Table 2). The most important parameter related to the sustainability of aquatic life, dissolved oxygen (DO), did not signi cantly differ among the study sites (p>0.05), with the maximum (6.3 mg/L) and minimum (5.25 mg/L) levels recorded at sites 5 and 6, respectively ( Table 2).
The minimum and maximum average surface water temperatures were 20.4°C and 21.4 º C recorded at sites 1 and 6, respectively ( Table 2). Electrical conductivity (EC, µS/cm) was not signi cantly different among the study sites (p>0.05) and varied from 202 to 239 (Table 2), with the maximum level occurring at site 6 where the river joins the Lake. TSS varied declining down the river, with the maximum (303.5) and minimum (231.1) levels occurring at sites 3 and 6, respectively.  Table 2). Except for pH, and temperature, levels of all measured physicochemical parameters were not signi cantly different across the study sites (Table 2).

Species composition, abundance and distribution of macrophytes
Sixteen macrophyte species belonged to fourteen families were identi ed, and their relative frequency and density are listed below in Table 1. Relatively, Cyperaceae and Poaceaewere dominant families both represented by 2 species each, while other families were represented by a single species. Among the identi ed macrophyte, 11 of them were emergent, while 3 were rooted with oating leaves (Nymphoides peltata and Nymphaea lotus) and 2 free oating (Pistia stratiotes and Azolla nilotica). Free oating macrophytes were shared the highest abundance followed by emergent. All the sites were dominated by emergent macrophytes that attained the highest relative frequency followed by the rooted emergent species, while free oating macrophyte species shared highest relative density followed by emergent species. The largest percentage compositions of macrophytes was comprised by the emergent group (66.58%), followed by rooted emergent (18.27%). The lowest percentage (15.17%) was contributed by free-oating species (Fig. 1). Azolla nilotica and Pistia stratiotes were shared the highest abundance and were the dominant macrophyte with the relative frequency of 7.24% and density of 40.91%, and 7.93% and26.54%, respectively. Azolla nilotica was occurred at all the sites, while Pistia stratiotes has occurred at site 3 only. Ludwigia stolonifera, Nymphoides peltata and Echinochloa stagnina followed in their dominancy with the relative frequency of 10%, 9.66% and 12.07%, and relative density of 9.13%, 7.03% and 4.93%, respectively. Even though Persicaria senegalensis did not dominant abundantly, it had high relative frequency (12.76%) and density (3.94%) ( Table 3). latifolia were present at site 3 only; the former species shared highest relative abundance next to Azolla nilotica, while the latter 3 species contributed the least to the relative abundance of macrophytes (Table  3). During the study period, site 3 shared the highest species richness (12), while site 1 and 2 shared 10 species at each site. Azolla nilotica was dominant throughout the sampling periods, while Pistia stratiotes was dominant throughout the sampling periods at site 3 only (Personal observation). Among the studied sites, site 3 shared the highest species richness (12), while sites 1 and 2 shared 10 species each. Down the course, the abundance of the macrophytes was increased and the highest abundance was recorded at site 3. Key community parameters (Shannon Diversity Index and evenness) generated for each site showed signi cant site-speci c variation (ANOVA, p<0.05). The overall macrophyte diversity index of site 3 was high (H′ = 1.44) and; the site had a value signi cantly higher than sites 1 and 2. Evenness also varied signi cantly among the sampling sites (p<0.05); site 3 (0.36) had value signi cantly higher than the site 1 and 2. Generally, site 3 had the highest species richness, abundance, Shannon diversity and evenness value.

Relationships between Macrophytes and Environmental Variables
Results of RDA showed that all of the environmental factors were the main determining factors that governed the distribution of macrophytes in Ketar River. The rst two axes explained 96.6% of the species-environment relation, while axis 1 only explained 81.1%. pH, temperature, conductivity and dissolved oxygen had signi cant positive with axis 1 and determined the distribution of Azolla nilotica, Nymphoide speltata and Ludwigia stolonifera and Pistia stratiotes while ammonium, total phosphorous and total suspended solid had a signi cant negative correlation with axis 1 and determined the distribution of most of the macrophytes. Axis II had signi cant positive correlation with nitrate and nitrite and predicted the distribution of Azollanilotica, Echinochloa stagnina, Ipomoea aquatic, Ludwigia stolonifera and Nymphoides peltata and ammonium had a signi cant negative correlation and predicted the distribution of most macrophytes. RDA supports the result obtained with ANOVA and the physicochemical parameters pH, temperature, conductivity and dissolved oxygen characterize at site3 (Table 6 and Fig. 2).

Physico-chemical parameters
The pH of Ketar River varied from 7.84 -8.11 indicating the alkaline nature of the river water, The pH values of the present study are within the range of desirable levels of pH (6.5-8.5) set by WHO [47] for optimal growth of aquatic organisms. The slight increase in pH observed along the river course (Table   2) may be associated with the deposition of sediment, which is known to contribute to an increase in pH Temperature is a factor of great importance for an aquatic ecosystem, as it affects the organisms, as well as the physical and chemical characteristics of water [49]. The present surface water temperatures are cooler than those reported by Degefu et al. [50] (23.53 -25.65 ºC) for Awash River. The lower level of surface water temperature of the present study might be due to the shading effect of macrophytes found along the banks of Ketar River, a condition, which was shown to impact river water temperature by, Lin and Herold [51] and Willis et al. [52]. The recorded mean levels of EC (202 to 239 µS/cm), which is a function of the amount of total dissolved salts [53], are lower than the permissible limit set by WHO [47] for drinking water. Koning and Roos [54] reported that the average EC value of typical, unpolluted rivers is approximately 350 mS/cm. Thus, the present result, which is less than 350 mS/cm indicates that the river water is suitable for direct domestic use. Compared to the levels of EC reported by Degefu et al. [50] for Awash River (327.67 -492.87 μS/cm), the present results for Ketar River indicate its much lower level of EC. This suggests that the river receives a low amount of dissolved inorganic substances in ionized form from its surface watershed [55].
The TSS content of water depends on the number of suspended particles, soil and silt, which are directly related to the turbidity of water. The present average values of TSS (231.1 -303.5 mg/L) are much higher than the permissible limit (150 mg/L) set by WHO [47] for drinking water. These high TSS values could be attributed to surface runoff and disposal of domestic sewage. Although the recorded TSS levels showed no signi cant differences among the study sites, the slightly higher values recorded at site 3 seem to have resulted from surface runoff from nearby agricultural lands. According to Akan et al. [56], river water with TSS values greater than 100 mg/L but less than 220 mg/L is classi ed as medium wastewater.
Thus, the overall mean TSS value for Ketar River is 267.3 mg/L, which warrants its classi cation as high wastewater.
Dissolved Oxygen is the most important parameter related to the sustainability of aquatic life. The lowest level of the range of concentrations of DO recorded in this study (5.25 -6.3 mg/L) occurred at sites 2 and 3, which receive agricultural runoff and animal wastes from nearby livestock holding operations. The high mean concentrations of DO recorded at the lower sites (sites 4 -6) could be due to the self puri cation of the water along the course of the river. The absence of statistically signi cant difference in the DO levels among sites at 95% con dence level ( Table 2) might be that the river owing down its course creates turbulence, which favors the dissolution of atmospheric oxygen [57].  Table 2). The values of nitrate are less than those reported by Degefu et al. [50] for Awash River, while those of ammonia are much higher than the levels documented by Degefu et al. [50]. In the present study, the concentration of both nitrate and ammonia were much lower than the  Among the identi ed macrophyte species, 11 of them were emergent, while 3 were rooted with oating leaves (Nymphoides peltata and Nymphaea lotus) and 2 free oating (Pistia stratiotes and Azolla nilotica). Ketar River is highly dominated by emergent macrophytes in terms of species diversity (compared with oating and submerged macrophytes), which could be due to their high tolerance of water-level uctuation [5] and water current [69]. Among the emergent species Ludwigia stolonifera, Echinochloa stagnina and Persicaria senegalensis were dominant and shared a relative frequency of 10%, 12.07% and 12.76%, relative density of 9.13%, 4.93% and 3.94%, respectively. However, emergent species did not show dominancy in abundance compared with free oating species.
On the contrary, free oating macrophytes shared the highest abundance. The free oating species of macrophytes identi ed in this study; A. nilotica (at all sites) and Pistia stratiotes (at site 3) shared highest abundance and were the dominant macrophytes throughout the sampling periods with the relative frequency of 7.24% and density of 40.91%, and 7.93% and 26.54%, respectively. A. nilotica was presented at all the study sites, while Pistia stratiotes presented at site 3 only. In contrast to the other macrophytes, A. nilotica has the ability to x nitrogen from the air [70] which could create favorable condition for it to compete with other macrophytes. P. stratiotes was also the dominant in abundance next to A. nilotica in the River. Temperature is one of the most important factors determining growth rates of free oating macrophytes and P. stratiotes can grow very quickly in tropical conditions [71]. The optimum temperature growth of Azolla ranges from 18 and 28°C [72]. Thus, in addition to a nutrient concentration in Ketar River, the temperature might be created favorable conditions for these dominant species (A. nilotica and P. stratiotes) to ourish and out-compete the other species. As Sadeghi et al. [73,74] reported, the presence of macrophytes communities also provides a good opportunity for the ourish of free oating macrophytes by breaking wind speed and water velocity.
Site 3, the site where faced minimal human impact was contributed the higher species diversity (12) and also con rmed the association of the distribution and growth of aquatic macrophytes with nutrient rich.
A. nilotica and P. stratiotes are the worst invasive oating macrophytes and have the ability to invade new habitats within a short period of time under a favorable environment. I observed that at the late dry season (before set on the wet season) the river was loaded with nutrients that could be encouraging the infestations of these macrophytes. In contrast to P. stratiotes, A. nilotica can exist even under low nutrient conditions by xing nitrogen from the air. The occurrence of A. nilotica in the River seems not to be affected by differences in the nutrient condition among sites, and its ability to colonize these varied sites indicates its potential to adapt to diverse trophic conditions. Additionally, the presence of an emergent macrophyte provides a good opportunity for Azolla to grow widely by breaking wind speed and water current.
Ketar River is the main tributary of Lake Ziway. Research conducted by and Tamire and Mengistou [21] indicates that Lake Ziway is dominated by emergent macrophytes. But, the littoral area of the Lake has been affected by anthropogenic activities; such as irrigation developments and abstraction of water for oriculture, and as result the water level of Lake Ziway has been declining [82, 83]. The above authors [82, 83] also reported that due to high evaporation, the lake showed a net loss of 74 million m 3 volume of water annually. The emergence of some invasive species of macrophyte such as P. stratiotes and A. nilotica including water hyacinth (Eichhornia crassipes) in Lake Ziway further makes worse the condition which calls for serious intervention in the catchment.

Conclusion
In general, the site (site 3) that was exposed to minimal human impact was rich in diversity and abundance of macrophytes. In this study, the emergent macrophytes were dominant in terms of species diversity, while free oating macrophytes (P. stratiotes and A. nilotica) were dominant in abundance. Biotic and abiotic factors lead to signi cant variations in distribution, diversity and abundance of aquatic macrophytes. In addition to good physical condition and presence of macrophte stand that provides a conducive environment, increases in nutrient loading from nearby create favorable conditions for the infestation of the invasive species (A. nilotica and P. stratiotes) to ourish and out-compete the other species of macrophyte. Therefore, anthropogenic activities that enhance nutrient addition to the River should be regulated.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. BMCEcologyandevolution.xlsx