Chapter 5 Discusion

5.1 Distribution and Abundance of Phytoplankton

The distribution and abundance of phytoplankton biomass throughout the year at the two investigated sites was observed to be highest on the nearshore waters and rapidly decreased further offshore. The results showed that the highest Chl-a concentration were observed during the month of April and August or during the rainy season than the dry season. The higher phytoplankton biomass in April and along the coast might be influenced by nutrients input from the land-based sources. During the rainy season (March through June), nutrients from land based sources are carried by the water to the ocean, and this would enrich the coastal waters and enhance productivity in these areas.

Higher phytoplankton biomass in August could be due to low sea surface temperature during the SE monsoon season that favours phytoplankton production. Furthermore, the higher phytoplankton biomass during this season could be associated with water mixing caused by strong winds that characterize the SE monsoon season. Mixing would bring up deep nutrient-rich waters to the surface, thereby stimulating phytoplankton production. McClanahan (1988) found that, phytoplankton biomass in the coastal waters of Tanzania is high during the SE monsoon season; and is mainly attributed by greater mixing of the water column, greater runoff and nutrient input from rivers and greater availability of biologically assimilable nitrogen.

There was a relatively higher phytoplankton biomass at the Zanzibar channel site than at the Pemba channel site (Figure 13). This might be attributed by high nitrate concentration in the Zanzibar channel (Tables 5 and 6) than for the Pemba channel site (Tables 7 and 8). The coast of Zanzibar channel receives sewage effluent from urban areas (Mohammed et al., 1993). Sewage is one of the rich nutrient sources to the ocean. Once in the ocean, the nutrients are utilized by the phytoplankton for photosynthesis to produce organic carbon into the ocean. The Pemba channel by contrast was associated with relatively low nutrient levels and this may attribute to the observed lower phytoplankton biomass on the Pemba channel. Suspended sediments and organic matter from the Pangani river at the Pemba channel site might also prevent light penetration for photosynthesis, thereby reducing phytoplankton production. This has been recorded as one of the major factors for the low productivity to many estuarine environments (Malone et al., 1988).

In assessing whether there was difference in biomass of phytoplankton between transects, it was found that there was no significant difference in phytoplankton biomass between the three transects during the dry season at both Pemba and Zanzibar channel sites (Table 3). Contrary to the dry season, statistically, phytoplankton biomass during the rainy season was not significantly different at the Zanzibar channel site while at the Pemba channel site there was a significant difference in phytoplankton biomass between the three transects (Table 3), with higher biomass at transect 2 compared to transects 1 and 3. Higher biomass at transect 2 at the Pemba channel site during the rainy season might be due to high nitrate and phosphate concentrations and low SST for this transect. When surface waters are cold, it is easier for deeper water to rise to the surface, bringing nutrient into the surface to sunlit areas where it is utilized by phytoplankton for productivity (McClanahan, 1988).

5.2 Correlation Between Chlorophyll-a and the Physico-chemical Variables

The Chl-a concentration obtained in this study was correlated with the physico-chemical variables and the results indicated that, nitrate and phosphate showed a positive relationship with Chl-a for both the dry and rainy seasons (Tables 9 and 11). However, nitrate showed a stronger relationship which was statistically significant compared to phosphate with a weak relationship. This might be due to the greater nitrogen requirement than phosphorus by the phytoplankton. Phytoplankton depend on nitrogen and phosphorus for making proteins, nucleic acids and other cell parts they need to survive, grow and reproduce. However they require 16 atoms of nitrogen, 1 atom of phosphorous and 106 atoms of carbon for making organic matter (Spokes, 2006).

Results showed that the increase of 1unit of phosphate is expected to increase the concentration of Chl-a in greater amount than a unit of nitrate, during the dry season (Table 10). Phosphate by itself had smaller impact in the abundance of phytoplankton, but in the multiple regression model the impact of phosphate increased because of the relationship existing between phosphate and other variables in the model.

However, lower nitrate and phosphate concentration was recorded during the rainy season when Chl-a concentration was higher. The lower nitrate and phosphate levels in the rainy season and the higher Chl-a concentration in this season is because; assimilable forms of nitrogen and phosphorus decrease simultaneously with the increase of phytoplankton biomass and the phytoplankton blooms had been recorded in the ocean when concentrations of phosphate and nitrate are at their lowest (Istvanovics et al., 1992).

Sea surface temperature showed a negative relationship with in situ and satellite Chl-a concentration in both correlation (Tables 9 and 11) and multiple regression analysis (Tables 10 and 12) during the dry and rainy seasons. This might be due to the influence of sea surface temperature on mixing processes in the ocean. When surface waters are cold, it is easier for deeper water to rise to the surface, bringing nutrients to sunlit areas where phytoplankton can use them and increase their biomass (McClanahan, 1988). When surface water is warm, cooler nutrient-rich water is trapped below; since the vertical layers of the ocean are not mixing, nutrients that have built up in deep waters cannot reach the surface (McClanahan, 1988). The result of low nutrients at the surface leads to low phytoplankton biomass. Mansor et al., (2001) and Solanki et al., (1998, 2003) also obtained similar results that SST and Chl-a were negatively related and then they used this information to model the fish distribution.

Dissolved oxygen showed a positive relationship with Chl-a concentration in both correlation and multiple regression analysis during dry and rainy seasons. Phytoplankton influence dissolved oxygen production and consumption (Connell et al., 1984). They produce oxygen during photosynthesis and this will increase DO concentration in the water column (Fukushima et al., 2004). Low phytoplankton biomass causes low oxygen production, hence low oxygen concentration into the ocean.

Salinity also showed a negative relationship with Chl-a concentration in both correlation and multiple regression analysis. Thus, low salinity was recorded at areas with high Chl-a concentration while high salinity was recorded at areas with low Chl-a concentration. Salinity decreases as freshwater from the river and rainfall runoff enters the ocean or when evaporation decreases in the ocean. When fresh water input increases, it sweeps nutrient from the land to the ocean as a result oceanic salinity decreases. The nutrient will be used by the phytoplankton to increase their biomass. In the western Indian Ocean, salinity had been recorded to be low during the rainy season when rainfall and discharge are higher (McClanahan, 1988). Consequently, phytoplankton biomass at this region is generally higher during this season and lower in the dry season (Bryceson, 1977).

In this study pH showed a positive correlation with Chl-a concentration in both dry and rainy seasons. Thus, low pH areas (acidic waters) were having low phytoplankton biomass while areas with high pH (alkaline waters) were having high phytoplankton biomass. The pH value is related to the concentration of phytoplankton biomass and abundance (Zhang et al., 2009). It is governed by the amount of carbon dioxide, which can react with water, as well as carbonate and bicarbonate to form complex but reversible carbonate systems (Wang et al., 2001). The carbon dioxide concentration is influenced by the phytoplankton, aquatic respiration, water temperature and oxidative decomposition of organic matter. Phytoplankton can transform carbon dioxide into organic matter and oxygen during photosynthesis, while aquatic respiration uses organic matter, resulting in the production of carbon dioxide (Zhang et al., 2009).

High phytoplankton abundance leads to increase in photosynthetic activities which will increase carbon dioxide uptake from the ocean, resulting in increase of pH (Wetzel 1983; Welch 1992). By contrast, the increase in carbon dioxide produced by aquatic respiration process and the low phytoplankton abundance causes a decrease in pH (Wetzel 1983; Welch 1992). However, high pH levels may have an inhibitory effect on phytoplankton metabolism via the carbonate system. At high pH levels, the availability of CO2 decreases and may become limiting to photosynthesis and growth of the phytoplankton (Zhang et al., 2009).

Also, the increase of CO2 give rise to ocean acidification and it is considered to have deleterious effects on calcifying organisms such as corals and some phytoplankton species such as the cocolithophorids. Some studies show that high acidic levels (low pH value) into the ocean cause calcium carbonate (which is the main skeletal material for many marine organisms as well as the phytoplankton) to dissolve (Pegler and Kempe, 1988). Therefore, pH mediated through CO2 may be an important abiotic factor affecting the ecology of marine phytoplankton.

In the multiple linear regression analysis, pH was found to be negatively related with Chl-a during the rainy season (Table 12). By itself (in correlation analysis) pH is positively correlated with Chl-a, but in the multiple linear regression model the increase in pH values leads to decrease in Chl-a concentration due to the relationship that exist between pH and other variables in the model.

5.3 Comparison Between in situ and Satellite Chlorophyll-a

Some of satellite images had no value due to cloud cover; this was a major challenge for matching up the satellite with the in situ data. The in situ measurement of surface Chl-a concentration was compared with ocean colour remote sensing images. The in situ Chl-a values varied between 0.001 and 0.770 mg/m3 with a mean value of 0.079 ± 0.021 mg/m3, whereas the mean Chl-a concentration of satellite images was 0.653 ± 0.037 mg/m3, with a maximum value of 1.691 mg/m3 (Table 17). These low values are typical of the oligotrophic waters as observed previously by other workers (Bryceson, 1977; Lugomela, 1996).

In order to quantify the major differences between the satellite Chl-a data and in situ Chl-a measurements the two datasets were correlated. The satellite Chl-a concentration showed strong positive correlation with the in situ Chl-a measurements at Zanzibar and Pemba channels during both the dry and rainy seasons. Considering the in situ Chl-a distribution (Figure 7 and 9), the relative pattern was the same as that of satellite Chl-a (Figure 6), and this revealed that there was a similar spatial pattern of Chl-a concentration and coastal waters showed to have a relatively higher concentration, which tend to decrease toward the open waters. Although the results showed that in situ Chl-a concentration and those extracted from MODIS sensor are generally in agreement, satellite Chl-a consistently over-estimated the in situ Chl-a.

Consistent over-estimation or under-estimation of Chl-a by MODIS sensor has also been reported by other workers (Joint and Groom, 2000; Doerffer and Schiller, 2007; Kampel et al., 2009). The discrepancy between the two datasets has often been attributed to either errors in the bio-optical algorithm performance, the satellite sensor calibration or in the atmospheric-correction algorithm (Joint and Groom, 2000; Doerffer and Schiller, 2007; Kampel et al., 2009).

The other factor which could account for the observed higher satellite Chl-a derived data relative to the in situ Chl-a measurements may be the Coloured Dissolved Organic Matter (CDOM) discharged by the rivers on both Zanzibar and Pemba channels. CDOM like phytoplankton also absorb blue light strongly, leading to higher Chl-a concentrations detected by the sensor (Sathyendranath et al., 2001). The difference between the in situ and satellite datasets may also be due to the time difference during data collection between the satellite and in situ measurements. This could be particularly significant when the chlorophyll-a distribution features are dynamic.