The "Three Amigos" pose in the Bahamas
Grahams Harbor is a windward, high-energy tropical lagoon located on the northeastern coast of San Salvador Island in the Bahamas (Figure 1). This shallow embayment, only 3m deep at low tide, is bounded on three sides by San Salvador Island to the south, North Point and Cut Cay to the east, and Catto Cay and the associated barrier reefs to the north (Gerace et al., 1999). While open waters exist to the west, Grahams Harbor is an embayed tropical lagoon that supports some the most abundant and diverse sea grass beds and algae communities on San Salvador Island (Gerace et al., 1999). Intricately connected to this tropical ecosystem is the presence of patch and fringing coral reefs that act as blockades against tidal surges and extreme water current velocities (Gerace et al, 1999). The presence of these fringing reefs not only protects Grahams Harbor against environmental extremes, but they also support coral and algal communities that are prevalent in such tropical marine environments. The correlation and symbiotic relationships between reefs, sea grasses, and algae in these kinds of shallow, marine embayments have been well researched, especially when subjected to nutrient fluxes from anthropogenic sources (McGlathery, 2001).
Figure 1. Geographic Setting of Grahams Harbor (Gerace et al., 1999)
Tropical marine embayments are among the most nutrient-sensitive environments in the world. Shallow tropical lagoons, such as Grahams Harbor, are extremely vulnerable to influxes of external nutrients, and the oligotrophic conditions present in such tropical lagoons are indicative of the delicate balance between nutrient cycling and floral communities (Reyes, 1991). While primary production does occur in these relatively cool, clear waters; green macro-algal blooms can exponentially increase when nutrient loading occurs from extraneous sources of nitrates and phosphates (Reyes, 1991). Thus, any influx of nutrients to an oligotrophic or mesotrophic habitat, such as Grahams Harbor, can promote eutrophic conditions within the localized ecosystem (Reyes, 1991).
Eutrophication occus when high nutrient concentrations of nutrients stimulate blooms of algae (e.g., phytoplankton). Resulting from an excess of nitrates and phosphates, eutrophication generally disrupts aquatic ecosystems by exacerbating primary production in the top of the water column and creating an oxygen deficiency near the bottom (Kinney and Roman, 1998). Although there are many associated effects, the major concern of eutrophication in seagrass-dominated estuaries is the increase in primary producing phytoplankton and algal blooms (Kinney and Roman, 1998). In delicate ecosystems where all available nutrients are consumed, any excess in nitrates or particulate-P will induce critical eutrophication (Reyes, 1991).
In Grahams Harbor, a potential source of nutrient loading exists from untreated wastewater generated at the Gerace Research Center, formerly the Bahamian Field Station. All wastewater created from the research center is discharged directly to Grahams Harbor via gravitational drainage. Presently, there is no secondary or tertiary treatment system, and there is no evidence to suggest that microbiological treatment has ever occurred (int., Glenn). Wastewater from the facility enters Grahams Harbor through a six (6) inch steel pipe that runs from the research center, flows under the beach, and discharges effluent approximately 25 feet into the embayment.
Untreated wastewater carries numerous aquatic pollutants that threaten shallow marine ecosystems. Among those pollutants are fecal bacteria, nutrients such as nitrates and particulate phosphate, and organic compounds such as ammonia that have detrimental effects on flora and fauna. Coelho et al. (2000) have demonstrated that sewage outfalls have a negative impact on sperm motility, fertilization and can cause increased mortality in seagrass germlings (Coelho et al, 2000). Furthermore, eutrophication can result in accelerated development of the early stages of some algal species, and nutrient enrichment can cause algal blooms and create an oxygen depletion in the littoral zone (Coelho et al, 2000). Other research suggests that eutrophication in marine waters can lead to a state of hypoxia that lowers dissolved oxygen, creates noxious algal blooms, and kills fish (Kaya, 1995). Because human interaction and usage is minimal in Grahams Harbor, the associated fecal pollution was not assessed in this study.
The purpose of this research was to determine if eutrophic conditions were present in Grahams Harbor. Based on field observations taken from 21 June 2003 to 25 June 2003, a team of researchers attempted collected and analyzed data to determine if eutrophic correlations existed between algal growth and proximity to a known sewage outfall.
We predicted that eutrophic conditions were present in the areas immediately adjacent to the sewage effluent pipe in Grahams Harbor. It was hypothesized that the highest percentage of algae abundance and seagrass coverage would be found in the areas nearest this outfall. We also predicted that the overall species diversity would be lowest immediately below the discharge and that one type of plant or algae would be thrive best in this area. Our third hypothesis was that seagrass height would be tallest in the immediate vicinity of the sewage discharge.
For this study, eutrophication was measured by assessing qualitative data that could be gathered and analyzed using the resources available at the Gerace Research Center. Initially, five transects were constructed in a 1000m2 area surrounding the sewage pipe. Transects were equally spaced 10 meters apart and ran perpendicular to the shore line with two (2) transects on each side of the pipe and one (1) transect extending directly from the pipe. Thus, a total of five (5) transects were constructed, with each transect extending 20m from the shoreline into the lagoon (Figure 2).
For each individual transect, sampling sites were predetermined to be taken at 4m intervals extending oceanward from the shoreline. Five (5) samples were taken along each 20m transect, with two samples collected at each point. Thus a grand total of 50 sampling sites were created. A plot-sampling technique was employed by using a 0.5m2 quadrant frame to collect our research. At each sampling site, the quadrant frame was randomly placed on the ocean bottom, and samples were collected. Again, two (2) samples, one on each side of the transect, were then collected for each 4m interval along the transect (Figure 2).
For each 0.5m2 sample plot, three (3) types of data were collected. These data were ™vegetation height, ‚percent coverage, and ńtotal number of species present. For the purposes of this study, emergent vegetation included both marine algae and seagrasses. Vegetation height was recorded using a metric tape, and the approximate mean vegetation height in each individual quadrant was recorded into a database. The percent vegetation coverage in each quadrant was estimated using visual inspection methods. The resultant percent coverage was classified into three categories: Low (0-33%), Medium (34-66%), and High (66-100%). Finally, the numbers of species present and individual species were determined by using taxonomic identification. In addition to these measurements, five salinity readings were also taken along the shoreline of each transect in attempt to detect the directional differences between water current and freshwater intimately associated with wastewater discharges.
Samples were analyzed using the Global F-test, or Analysis of Variance (ANOVA), to detect differences between each measurement upstream of the effluent pipe and downstream of the pipe. It was determined that a P-value of 0.05, or 95% confident coefficient, would be significant enough to analyze these data. Since the results of the ANOVA model can only detect whether or not a difference exists, it was necessary to conduct a post-hoc evaluation to determine exactly which samples were significantly different. The Scheffe post-hoc test was chosen for its relative conservative ability to detect significant differences. Using this pair-wise comparison test, significant differences between each transect could be detected, and generalizations could easily be made between samples and their relative proximity to the effluent sewage pipe. All statistical differences were detecting using a Macintosh version of SuperANOVA.
From the ANOVA model constructed for emergent vegetation height, the results indicated that a statistically significant difference existed between the five transects. With a p-value of 0.0006, we can be 95% confidence that a difference in mean vegetation height exists between the five transects (Figure 3). When graphed, Figure 3 shows that an inverted bell-shaped distribution results when the mean vegetation height is compared among the five transects. From these results it appears that mean vegetation height increases exponentially (over 150%) in the transects furthest from the pipe. Interestingly, the mean vegetation height appears to increase in the same proportions on each side of the pipe, i.e. at both 10m and 20m transects upstream and down (Figure 3).
From the ANOVA model constructed to analyze percent cover, the resultant p-value of 0.0001 indicates that a significant difference exists between the transects and the percent of vegetation coverage in those transects (Figure 4). With 95% confidence, Figure 4 shows that this difference is again detectable with transect C having the lowest mean percent cover. Unlike the distribution in the graph portraying vegetation height, the graph portraying percent coverage indicates that higher percent coverage exists in transects A and B, which are both located downstream of the effluent pipe (Figure 4).
The Scheffes post hoc analysis indicates that significant differences lie between transects A and C, transects A and D, transects A and B, and transects A and E (Table 2). Therefore, with 95% confidence, these results indicate that transect A, which was 20m downstream of the pipe, has a significantly higher percent coverage of vegetation as compared to all other transects (Table 2). At the 0.05 confidence level, no other significant differences were detected between the other four transects (Table 2).
From the ANOVA model constructed to analyze species diversity, the resultant p-value of 0.0014 indicates that a significant difference exists between the transects and the number of species present in those transects (Figure 5). Similar to the vegetation height model, this ANOVA model indicates that a difference is detectable and the distribution represents an inverted bell-shape. From the box plot portraying number of species, it appears that mean species diversity increases exponentially (over 150%) in the transects furthest from the pipe (Figure 5).
Again using the Scheffe’s pair-wise comparison test, a post hoc evaluation of the ANVOA model was conducted to determine the specific differences between transects. From this analysis, it was discovered that, with 95% confidence, the mean number of species in plots from transect C were statistically different from those in transects A and E, which are 20m upstream and downstream, respectively (Table 3). It appears that the transects furthest from the pipe contain almost twice as many species as the transect perpendicular to the effluent wastewater pipe. All other transects were found to have similar mean vegetation heights at the 0.05 level of confidence (Table 3)
Based on the findings in this research, there is some evidence to suggest that eutrophic conditions may be present near the sewage effluent pipe in Grahams Harbor. However, only one of our hypotheses was supported based on these results. Our results demonstrated that the relative proximity to effluent discharges are only part of the equation when trying to determine if eutrophic conditions exist in a unconfined system, such as a tropical lagoon. Numerous factors must be taken into account to definitively assess eutrophication in marine ecosystems (Menon et al, 2003). Some of these such other factors include available sunlight, antecedent dry weather periods, prevailing currents, seasonal differences, and relative speed of nutrient mixing in the littoral zone (Menon et. al, 2003). Such factors were not considered in this study, and only emergent vegetation was used as a bioindicator for eutrophication.
Of our three hypotheses, only one was supported by the research. The number of species found along each transect appeared to increase linearly with distance from the discharge pipe; thus, our hypothesis that the species diversity would be lowest near the pipe was supported. In fact, one species (Thalassia testudinum) seemed to dominate in this region and push out all other competition. This could be due to the thermal increases associated with wastewater or the fact than an artificial reef, comprised of abandoned pipes, was located near the outfall.
Our hypothesis that the areas nearest the sewage would have the highest percent coverage was not supported by the data. In fact, the opposite effect was observed, and the mean percent coverage was lowest near the pipe and highest in the transect located 20m downstream of the discharge pipe. If indicative of eutrophic conditions, this effect suggests that prevailing water currents in the easterly directly play an important role in nutrient transport.
Similarly, the third hypothesis was rejected because the tallest vegetation was not found nearest the pipe, as we had predicted; rather, mean vegetation height was greatest in the two transects furthest from the effluent. As the results indicate, mean vegetation height was similar for both transects located 20m from the effluent. This phenomenon (inverted bell-shape distribution) is quite interesting because it may suggest that water currents are flowing in both directions at different times in the embayment, and prevailing water currents may be flowing in both directions parallel to the shoreline.
Although two of our hypotheses were invalidated, there is significant evidence that suggests eutrophic conditions may still be present in Grahams Harbor, just not in the areas we expected. The research indicated that nutrients are transported with the current and eutrophic processes occur downstream of the initial discharge point. This observation should be taken with caution, however, because there are many other factors which have been described earlier. In addition, it should be noted that the decline in species, percent coverage, and vegetation height may also be a result of the presence of a small artificial reef that was located near the mouth of the effluent pipe.
Future research should consider quantitative data such as total biomass within each sample quadrant as well as an assessment of ambient factors that could be acting on the balance of nutrient cycling in Grahams Harbor. Shallow tropical embayments, which are classified as oligtrophic habitats are some of the most nutrient sensitive areas in the world. The normal nutrient cycling in such areas is delicately balanced. Therefore, chemical data such as nitrate, ammonia, and particulate phosphorous concentrations should be taken in this sample area and compared to a baseline reading taken from further upstream.
Coleho, S., Rijstebil, J., and Brown, M. 2000. Impacts of anthropogenic effect on the early
development stages of seaweeds. Journal of Aquatic Ecosystem Stress and Recovery. v. 7, no.1, pg. 317-333.
Gerace, Donald T., Ostrander, G.K. and Smith, G.W. 1999. San Salvador Island, Bahamas. CSI:
Coastal Regions and Small Islands Paper Series. Accessed from the United Nations Educational, Scientific, and Cultural Organization (UNESCO) website on 31 July 2003 at: http://www.unesco.org/csi/pub/papers/gerace.htm.
Kaya, Yusuf. 1995. Some environmental aspects with marine disposal systems with particular
reference to UK waters. Water Science and Technology. v. 32, no. 2, pg. 167-174.
Kinney, E.H and Roman, C.T. 1998. Response of primary producers to nutrient enrichment in a
shallow estuary. Marine Ecology Progress Series. v.163, pg. 89-98
Menon, H.H, Balchand, A.N., and Menon N.R. 2000. Hydrobiology of the Cochin tropical
backwater system. Hydrobiologia. v. 430, pg. 149-183.
McGlathery, Karen J. 2001. Macroalgal blooms contribute to the decline of seagrass in nutrient
enriched coastal waters. Journal of Phycology. v.37, no.1, pg. 453-56.
Reyes, Merino M. 1991. Dissolved oxygen dynamics and eutrophication in a shallow, well-
mixed tropical lagoon. Estuaries. v. 14, no.4, pg 372-81.
Valiela, I., McClelland, J., Harwell, J., and Behr, P.J. 1997. Macroalgal blooms in shallow
estuaries. Limnology and Oceanography. v. 42, pg. 1105-
Return to Topic Menu
We also have a GUIDE for depositing articles, images, data, etc in your research folders.
Article complete. Click HERE to return to the Pre-Course Presentation Outline and Paper Posting Menu. Or, you can return to the course syllabus
WEATHER & EARTH SCIENCE RESOURCES
OTHER ACADEMIC COURSES, STUDENT RESEARCH, OTHER STUFF
TEACHING TOOLS & OTHER STUFF
It is 5:14:12 AM on Wednesday, May 23, 2018. Last Update: Wednesday, May 7, 2014