Discussion

The results from the monthly precipitation data indicate that while the difference between maximum and minimum rainfall in the wet season is significantly higher, both sets of data are relatively normally distributed around the mean. Each set of data contains one month that is more than one standard deviation from the mean, suggesting that rainfall patterns are fairly consistent per season. This hypothesis, however, is limited by a lack of sufficient data, and thus, definitive conclusions regarding consistent rainfall patterns cannot be made. This is also due to the fact that rainfall is highly affected by: global climate change, which is presumed to enhance monthly precipitation levels in the wet season as well as the increasing occurrence of El Niño events, which has been shown to cause drought conditions, both of which are not accounted for in this study (Lasco et al., 2010). Despite this lack of data, the monthly precipitation data presents implications for trends in streamflow and discharge from rivers draining into the bay. According to Lasco et al. (2010), rainfall, among various other processes including land use and land cover, soil, and topography, is a key factor that influences streamflow and discharge from rivers. Lasco et al. (2010) observed that between 1980 and 2000, observed streamflow data reveal that as rainfall rises and declines, due to variations between the wet and dry seasons, streamflow does as well, which can be seen in Figure 6. The monthly streamflow averages for the wet and the dry season are very similar to the pattern of rainfall over this period.

Figure 6: Observed monthly average streamflow in relation to monthly average rainfall (taken from Lasco et al., 2010).

According to Hallegraeff (1989), increased discharge of industrial and human wastes into major waterways dramatically enhances nutrient concentrations and sediment loads transported in runoff to coastal waters when rainfall levels peak during the wet season. Villanoy et al. (2006) shows that the largest freshwater discharge originates from rivers along the head, or upper region, of the bay, where the Pampanga River is located. While nutrient concentrations cannot be assessed or examined in this particular study, Hallegraeff (1989) reveals that the excess nutrients entering Manila Bay as a result of increased rainfall and subsequent runoff from the Pampanga River now enable sufficient reproduction of dinoflagellates to overcome organism dilution losses, which significantly augments the potential for successful alongshore emigration (Hallegraeff, 1989). Therefore, with these favorable conditions and enhanced rates of reproduction during the wet season, it can be expected that there will be higher concentrations of algae in the water, and thus higher levels of greenness than during the dry season, which has significantly less rainfall (mean dry season: 15mm, mean wet season: 352.93mm). The results from the aforementioned studies provide evidence that substantiates the strong relationship found (r value: .84 for dry season, r value: .79 for wet season) between the two variables explored in this study including monthly precipitation levels and the band 2 reflectance levels (indicative of algae) in the water column.

Due to the fact that both data sets were very small, each season only containing 4 observations, correlation coefficient values tend to be easily biased by outliers in the data set, which is most likely what happened in the case of October 2007 that had a considerably lower precipitation level, yet fairly similar cursor value, to the other months. A probable explanation for the significantly lower precipitation value of this month can be attributed to the fact that while October is still part of the wet (southwest monsoon) season, it is approaching the end of it and consistently has much lower rainfall levels than the months of July  - September (Lasco et al., 2010). The fact that the cursor value remained relatively similar, despite decreased precipitation levels in comparison to the other months, may be due to the fact that the July-September months of 2007 had abnormally high precipitation levels, and the high concentration of algae in the water in October 2007 may be residual from the previous months. It is interesting to note that while the cursor values used for the scatter plot and analysis were taken from the same coordinates in every image near the mouth of the left branch of the Pampanga River, the highest cursor value for each image in any location (presumably where the clouds were brightest) was also noted, and October 2007 had the highest value of 126 out of all the wet season months.

The drastic differences between cursor values from the wet and the dry season, with values from the wet season tripling those from the dry season, provide evidence for extreme variations in algae formation, distribution, and concentration as a result of seasonality. Overall, when looking at the satellite images from the dry season, they appear to be relatively similar, with very minor differences in the extent of the plume. The plume, observed by the cloudy regions of the bay, appears to spread throughout the region more so in February 1989 than in other years; however, the algae concentration is still very low and is thus unlikely to be harmful to marine organisms or reef ecosystems. Moreover, further authenticating the unlikelihood of the bloom’s possible harmful nature, Azanza and Miranda (2001) show that P. bahamense blooms consistently start at the onset of the rainy season just after a dry period, which is crucial for the initiation of these toxic blooms. Thus, while January 2001 has the highest cursor value as well as the brightest magenta clouds in comparison to the other dry season months, it could be attributed to the higher precipitation level than the other months and, and the bloom is not presumed to be harmful to the coastal ecosystem. July 2011 has the highest precipitation level, which is typical of that month in this region. It could also be to global climate change, which is hypothesized to increase precipitation levels during the wet season (Lasco et al., 2010); however, the exact reason cannot be assessed at this time.

As previously stated in this study and supported by Hallegraeff (1989), the initiation of these potentially harmful blooms and the potential for successful transport in Manila Bay is highly contingent upon two processes. The first is favorable environmental conditions, which become present with the onset of the rainy season as nutrient levels are enhanced and delivered into coastal waters along the transport path resulting in high dinoflagellate concentrations. This process is explored in this study and the drastic differences observed between cursor values, and thus algae concentration, in the wet and the dry seasons is extremely apparent. Favorable environmental conditions during the rainy season alone, however, do not explain why the satellite images from the wet season reveal such extreme variations (especially compared to the dry season) in the offshore extent and distribution of the algal blooms. The second process described by Hallegraeff (1989) in terms of what augments successful transport and reproduction of the dinoflagellates, is attributed to the presence of persistent wind-driven currents in the proper direction and at a time when the dinoflagellates have bloomed and formed into surface patches. The combination of these two processes results in the initial settlement and formation of the blooms and the eventual transport and reproduction of dinoflagellates that have the potential to cause harm to the coastal and marine ecosystem (especially when dealing with toxic P. bahamense which is the species found in Manila Bay). Moreover, when the conditions are not quite as favorable, due P. bahamense ability to form resting cyst beds, this toxic dinoflagellate still remains in the sediment and becomes resuspended with the proper wind-driven currents that occur during the southwest monsoon. Although this study is limited in regards to lack of data showing wind current patterns and their effects, this second process can provide an explanation as to why there are such noticeable variations observed in the wet season images in regards to offshore extent and distribution.

The northeast monsoon, typically occurring during the dry season months of December to March, achieves maximum strength occurring in January and weakens in March (Lasco et al., 2010). This could explain why the algae concentrations observed in the image from January 1996, when compared to the March image from the same year, extended significantly further down the western coast. Stronger winds in January may have augmented alongshore transport; however, the algae was in very low concentrations and not harmful at this time. Villanoy et al. (2006) hypothesized that during the northeast monsoon (period of unfavorable environmental conditions), the bottom currents may not be strong enough to resuspend cysts. Moreover, high turbulence in the water, characteristic of the NE monsoon, results in dilution losses for the toxic dinoflagellate, which can explain why the band 2 values in the dry season were significantly lower than during the wet season. Villanoy et al. (2006) explains that during cooler months, convective mixing and high turbulence results in a weakly stratified water column, which is in accordance with our results from the January 1996 thermal image that certainly reflects high turbulence and poor uniformity.

Conversely, strong wind forcing, which consequently results in an increase in bottom current velocity during the SW monsoon, is crucial for the resuspension of cysts throughout the water column (Villanoy et al., 2006). The blooms are perpetuated and can counteract dilution losses that occur during the NE monsoon because of the stable subsurface water and subsequent decreased vertical mixing during the SW monsoon (Villanoy et al., 2006).

In the August 2003 and July 2011 images, the bloom extends significantly greater distances down the western coastline and into the bay than in the other wet season months. Villanoy et al. (2006) explains that the dispersal distances of the blooms increase if the P. bahamense cells are found higher in the water column. Although this cannot be seen from the satellite images, this could explain why the blooms in August 2003 and July 2011 extend further distances. Moreover, as highlighted throughout the results, blooms were consistently observed in the highest concentrations along the northern coastline of the Pampanga region. While this certainly can be attributed to the drainage of the Pampanga River into coastal waters, Villanoy et al. (2006) states that the formation of cysts after a bloom is adjacent to the northwest (Pampanga) for blooms originating from the west and the north. Because the direction of advection is almost always along wind direction (Villanoy et al. 2006), blooms forming in the west and the north will be advected along the west-northwest coast, in accordance with the results from this study. Due to the fact that nutrient levels will almost always be highest nearshore due to coastal activities and rivers, this could explain why the strongest concentrations of algae were observed nearshore.  

Improvements for this study could include observations from not only the Pampanga/Bulacan region, but also from the east coast of the bay where the Pasig River clearly plays a role in bloom formation via runoff and increased sedimentation. All four images from the wet season show bloom formation near the mouth of this river and by including this in the study, a more comprehensive understanding of bloom dynamics of the entire bay could be possible. Finally, since Manila Bay is in the center of the Philippines and a booming trade center, coral reefs are relatively nonexistent. Future research on how these algal blooms specifically and adversely affects coral reefs would also be important.