Water Quality of Estuarine Ponds in Hilo, Hawaii: Monitoring of Nutrients, Benthic Organisms and Bacteria

Authors:  Kathy L. Seiber(1), Shauna K. Tom(1), Toni M. Gregg(1), Bellia Rivera-Poy (2), and Misaki Takabayashi(1) 
Institution:  1)University of Hawaii at Hilo, Marine Science Department 200 W. Kawili St, Hilo HI 96720; 2)Hawaii Department of Health, 191 Kuawa St, Hilo HI 96720
Date:  December 2006


The coastal area of Hilo, Hawaii are dotted with estuarine ponds with varying degrees of influence by input of terrigenous water by run-off and ground water. These estuarine ponds are popular for recreational uses such as swimming and collection of benthic diatom, Melosira sp. as fishing bait. The purpose of our study is to determine if the ponds act as a buffer zone for the adjacent coral formations and to generate a baseline data set of the water quality of these ponds and compare the bacterial concentrations to the Hawaii Department of Health (HDOH) and national Environmental Protection Agency standards. Nutrient levels, Clostridium perfringens and Enterococcus faecalis concentrations within the ponds were monitored and surveys of benthic macroalgae and invertebrates were conducted monthly from September 2005 to March 2006. Nutrient concentrations varied spatially and temporally. Benthic flora and fauna cover and assemblage varied temporarily within ponds and spatially between ponds. The microbe levels also showed high frequency and range of concentrations. The concentrations of C. perfringens and E. faecalis were consistently higher than the state standards some as much as eight times as high. Based on this study the estuarine ponds do not appear to be acting as a buffer zone between the anthropogenic factors and coral formations. The water quality of these ponds is affected by many anthropogenic and natural factors and varies greatly temporally and spatially.


Along the Keaukaha coast of Hilo, Hawaii, there are several estuarine ponds. These ponds are heavily used for cultural and recreational purposes. This area of the coast is believed to be strongly influenced by natural and anthropogenic runoff, freshwater springs, and ocean mixing. Fringing coral reef formations are found adjacent to these estuarine ponds. Since coral reefs cannot tolerate high levels of nutrients and freshwater, how are these reefs found so close to areas that contain those conditions? These estuarine ponds may be acting as a buffer zone between the high nutrient, freshwater areas and the coastal fringing coral reefs.

Since these ponds are heavily used and influenced by anthropogenic factors, monitoring of the nutrient levels, benthic organisms and microbial community is crucial. All these characteristics are great indicators of water quality and should be monitored regularly (National Resources Defense Council, 2004). These characteristics should be monitored because they provide information on the overall health of the area.

According to the National Resources Defense Council (2004), nutrients are a major cause of estuary degradation. Excess nutrients can lead to algal blooms and can result in poor water quality such as anoxia (Shayler et al. 2000). When nitrogen levels are exceedingly high in the estuarine ponds, it poses a threat to the aquatic ecosystem and to humans who use the ponds for recreation (Hsieh 2000). An example of how nutrients enter the ecosystem is through point sources such as leaching, groundwater runoff, and sewer and waste water overflow. In areas where coral reefs are found it is important to monitor nutrients because corals are adapted to naturally oligotrophic (i.e. low in inorganic nutrient concentrations) waters, many reef organisms, including corals, are highly sensitive to disturbances in nutrient cycling which are primarily from terrigenous input (Fabricious 2005). Some effects of excessive inorganic nutrients include: inhibited coral growth leading to increases of algae and benthic organism abundance (Kinsey and Davies 1979); increases in photosynthesis by zooxanthellae causing increased removal of dissolved inorganic carbon (Ferrier-Pages et al. 2004); and decreases in coral fecundity (Cox and Ward 2002 and Loya et al. 2004).

Benthic organism and substrate composition is also an excellent indicator of water quality (Natural Resource Defense Council, 2004). If there are changes in the organisms or substrate covers, it could indicate changes in water conditions and nutrients. The substrate composition can be an excellent indicator of nutrient levels. If there are high levels of nutrients then algae blooms are likely (Shayler et al. 2000). The biota is also likely to provide take up nutrients and facilitate a buffering effect.

Microbe species have been used extensively worldwide to test water quality and for fecal/sewage contamination of the environment (Shibata et al. 2004). Monitoring the amount the pathogenic bacteria is highly recommended at recreational sites including beaches and the estuarine ponds in this study. The Environmental Protection Agency (EPA) recommends using the following types of bacteria as indicator species: total coliform, fecal coliform, Esherichia coli and Enterococcus faecalis. These bacteria are not only indicators of sewage, but are also pathogenic (Petit et al. 1999 and Sasaki et al. 2000).

The species recommended by the EPA can thrive in tropical waters. However, E. coli and E. faecalis also occur naturally in Hawaiian soil (NRDC 2004). This means that if elevated levels of E. coli and E. faecalis are found in the water, it may not be exclusively indicative of sewage pollution (Shibata et al. 2004). Due to the presence of soil borne Enterococcus sp., the state of Hawaii has adopted a unique standard for water quality; the state uses a lower enterococci standard compared to the nation as well as another species, Clostridium perfringens, to monitor water quality (NRDC 2004, Shibata et al. 2004). Unlike other indicators, C. perfringens does not multiply in aerobic environments, but it can persist for long periods in water. Since it can survive for extended periods, in situ counts of C. perfringens may not indicate recent contamination (Shibata et al. 2004). Bacteria density is measured by the number of colony forming units (CFU) which are the visible colonies of a pure culture of the bacteria. The state CFU standard is for C. perfringens is five CFU and the state standard for E. faecalis is seven CFU with a federal standard of 35 CFU per water sample (NRDC 2004). According to the National Resource Defense Council (2004), Hawaii has one of the strictest bacteria standards. By monitoring multiple species, Hawaii is able to draw a better picture of the water quality related to microbes.

Water quality of the estuarine ponds in Hilo, Hawaii has not been monitored extensively or consistently. The current survey is part of a continuing baseline study to examine the water quality, interactions of bacteria and nutrient concentrations, and the ecological importance of three coastal estuarine ponds to the adjacent coral reef ecosystem. The three goals of this study are to (1) monitor nutrient and bacterial concentrations; (2) monitor the abundance of benthic organisms and if they are responding to nutrient concentrations; and (3) determine if the ponds are reducing the nutrient load before estuarine water mixes with oceanic water.


Site Description

The three estuarine ponds that were surveyed are located along the Keaukaha coast in Hilo, Hawaii (Figure 1a).





Ice Pond (19'72", 155'06") is approximately 4,000m2 in area and 7m at the deepest point during high tide. It is the first pond at the north end of Keaukaha bordered directly by two roads and a restaurant. Wave action is low due to the Hilo Bay breakwall, but this pond is tidally influenced. Lalakea pond (19'73", 155'02") is approximately 3,200m2 in area and 2m at the deepest point during high tide. Lalakea has a small channel that connects the pond to the open ocean and is influenced by tides as well as waves. Leleiwi pond (19'73", 155'61") is approximately 3,000m2 and 2.5m at the deepest point during high tide. Of the ponds studied, Leleiwi has the most connectivity to the ocean allowing for lots of wave action and mixing. All ponds have ground water input of unknown volumes, and Ice Pond is suspected to receive more runoff than the others due to bordering roads (Figure 1b).

Nutrient Analysis

Nutrient water samples were collected every month from all three sites from October 2005 to March 2006, except during December 2005 using sterile 50 ml tubes. Water samples were collected from the surface and bottom of the ponds and the oceanic water adjacent to the ponds at low and high tides during the same tidal cycle during the monthly sampling. Temperature, dissolved oxygen concentration and salinity of water samples were measured in the field (YSI 85, Fondriest Environmental, Ohio). The samples were frozen (-20°C) for one to four weeks until the nutrient analysis of nitrate/nitrite, ammonia and phosphate were measured without filtration (Pulse Autoanalyzer, Technicon).

Benthic Survey

Three sites in each of the estuarine ponds were randomly selected for repeated monthly surveys of benthic organisms from September 2005 to March 2006 except December 2005. Percent cover of benthic organisms was analyzed using five random points on a 25 square, 1m x 1m quadrat.

Bacterium Concentrations

Water samples for bacteria analysis were collected from the three estuarine ponds once every month using sterilized 1L bottles. Two bottles were used to collect surface water samples and the other two were used to collect bottom water samples. One surface water and one bottom water sample were given to Hawaii Department of Health laboratory for microbe quantification; the other surface and bottom sample were used to record temperature, dissolved oxygen and salinity using a YSI 85 in the field. 100 ml of these water samples were filtered through a sterile 47 mm membrane filter (HCWG, Millipore Corp or GN-6, Gelman Sciences). The filters are then placed on to mCP Agar plates, selected for C. perfringens and incubated anaerobically at 45o C ± 0.5o C for 18-24 hours. After incubation, the yellow colonies were marked using a small fluorescent lamp to allow for best visibility. The colonies were then exposed to concentrated ammonium hydroxide for 20-30 seconds, to stain red. The colonies were then counted and recorded using the following formula: (C. perfringens/ 100 ml = (No. of CFU/ volume of sample filtered) x 100.

Enterolert (IDEXX, Maine) was used to quantify the E. faecalis density in the same water samples as for C. perfringens. The Quanti-Tray Enumeration Procedure (IDEXX, Maine) was used in which water samples were mixed and placed into a Quanti-Tray. The Quanti-Tray was then sealed and placed into an incubator at 410 C ± 0.5o C for 24 hours.

Statistical Analysis

All statistical analysis was conducted in Minitab (Minitab Inc. Pennsylvania). ANOVA was used to test the nutrient concentrations that were found to have normal distributions versus date, site, area (inside or outside) of pond, and depth (surface or bottom). The data that was not normally distributed were analyzed with the Kruskall-Wallace analysis. Tukey's test was used for post-hoc analysis of factors that were found to be statistically significant in all analysis. Correlations of rainfall to nutrient and bacteria concentrations were also done in Minitab.


Nutrient Analysis

The concentrations of all the nutrients showed spatial and temporal variability (p less than 0.01). Nitrate/nitrite concentrations varied from 1.37 to 36.0µM, mean 19.5 (±10.5). The nitrate/nitrite concentrations were highest in January at Lalakea (39.4 ± 2.94 µM, Figure 2a). Ammonia concentrations varied from 0.00 to 2.54µM and were highest at Leleiwi during October and March (2.30 ± 1.49 µM and 2.54 ± 1.11 µM respectively, Figure 2b).





Phosphate concentrations varied from 0.00 to 1.43µM and the highest concentrations were at Ice Pond especially during March (1.43 ± 0.156µM). Each pond had a specific nutrient that was found in the highest quantities; for example nitrate/nitrite was found in the highest amounts at Lalakea; ammonia was found in the highest amounts in Leleiwi; and phosphate was found in the highest amounts at Ice Pond (Figures 2a-2c).

Depth was found to be significant for only ammonia concentrations (p less than 0.01) with a surface mean of 0.50µM (± 0.68) and a bottom mean of 0.87µM (± 1.15). Area, inside or outside of the ponds, was also found to be statistically significant for only ammonia (p less than 0.05, 0.60 ± 0.86 µM and 0.77 ± 1.14 µM respectively).

The nutrient concentrations did not have strong correlations with the amount of rainfall but there were statistically significant correlations (Table 1 and Figures 3a-3f). The correlation of each nutrient with rain during the previous 24 hours and seven days had Pearson correlation coefficient values close around 0.6. There was not a drastic difference in the correlation values between rainfall during the previous 24 hours and rainfall during the previous seven days. Ammonia had lower correlation values than the other nutrients but still had statistically significant correlations.



Benthic Survey

The filamentous diatom, Melosira sp. and Nerita picea, black nerite, were the major macroscopic benthic organisms found in all three ponds (Figures 4a-4c).



Other organisms found in the ponds were Neritina vespertina, Brachidontes crebristriarus as well as green, red, brown and crustose coralline algae (Figures 4a-4c). At Ice Pond Melosira sp. was the dominant organism with N. picea, N. vespertina and crustose coralline algae being minor constituents (Figure 4a). Lalakea was found to contain a variety of organisms such as Melosira sp., N. picea, green and brown algae (Figure 4b). Leleiwi also contained the same variety of organisms as Lalakea but not brown algae. Also present at Leleiwi were B. crebristriarus, red algae and crustose coralline algae (Figure 4c). From correlation analysis between nutrients and the percent cover of Melosira sp. did not show statistical significance benthic plants do not show percent cover variation with nutrient concentration variation. There was also no herbivore effects found on the percent cover of Melosira sp.

Bacteria Abundance Clostridium perfringens and Enterococcus faecalis density varied spatially (p less than 0.01). Statistical correlations of each bacteria species with precipitation within seven hours and 24 hours of sampling showed no significance for each bacteria and rainfall. The highest amount C. perfringens was at Leleiwi during November (40.5 ± 13.44 CFU, Figure 5a). The state CFU standard is five CFU; the abundance of C. perfringens in the collected water samples is as high as eight times the standard.



Lalakea had the highest CFU of E. faecalis with the highest CFU during October (180 ± 113.12 CFU, Figure 5b). The CFU standard for E. faecalis in Hawaii is seven CFU per water sample and the federal standard is 35 CFU per water sample. The density of E. faecalis in the water samples was always much higher than the state and federal standards.


The anthropogenic effects on costal water quality are of major public concern in areas such as the Keaukaha coast in Hilo, Hawaii. These areas are used for recreation and cultural practices. The ponds studied are in highly populated areas are extremely susceptible to anthropogenic factors.

In this area the nutrient concentrations in the estuarine and coastal waters might be related to several unregulated sources of inorganic nutrients. Hawaii heavily relies on septic tanks and cesspools. In 2005 it was estimated that there were 100,000 cesspools (Friedlander et al. 2005), some of which are found adjacent to these estuarine ponds and the ocean. The septic tanks and cesspools leak nutrients and many other things such as bacteria into the water (Bose and Gerald 2006). The Island of Hawaii also still has many farms which may use fertilizers which release excess nutrients into the water as well. In Hawaii, it is important to monitor nutrient concentrations and their impacts on the fringing coral reefs. There have been several studies conducted on the effects of nutrients on coral reefs and the reefs in Hawaii could be experiencing some of those effects.

Lalakea had two of the highest measured concentrations of nitrate/nitrite. Nitrate/nitrite has been found to be leeched from cesspools (Bose and Gerald 2006). There are many houses that are adjacent to the pond and these houses are believed to have cesspools like many of the other homes in the area. If these homes do have cesspools they may be contributing to the higher amount of nitrate/nitrite at Lalakea. Ammonia concentrations were the highest at Leleiwi. The exact source of the high ammonia is not known however, there is a restroom located adjacent to the pond and probably uses a cesspool system which would also leak ammonia. Ammonia can also be reduced biologically or chemically from nitrate which would result in higher concentrations. Ice Pond had the highest concentrations of phosphate and also had the highest percent cover of the Melosira sp. diatom which may be related to the amount of phosphorus present. The rainfall within seven days of sampling had overall stronger correlation coefficients leading us to believe that rainfall during the previous week has more of an influence on nutrient concentrations that rainfall within 24 hours.

The analysis of the benthic communities showed that while some organisms are more dominant than others in each pond the abundance of organisms in the ponds are still patchy. However, some patterns were seen among the organisms and the ponds. Ponds that were more connected to the open ocean, Lalakea and Leleiwi, had higher species diversity than the pond with less connectivity, Ice Pond. The higher connectivity with oceanic water may facilitate more supplies of organisms that inhabit the ponds. The benthic organisms do not appear to be having a measurable affect or response to the nutrient concentrations in each pond or vice versa. Melosira sp., the species believed to have the most effect on nutrient concentration, did not show any correlation with nutrient concentrations; the overall buffering action of these ponds related to how much the flora is reducing the inorganic nutrients from the terrigenous water appears to be low. The organisms in the ponds are not playing a role in the buffering as expected which could be due to the patchy nature of the organisms. The high connectivity of the study sites to the adjacent ocean which could enhance mixing and lead to an appeared reduced buffering affect from the ponds. The pond with the least connectivity to the ocean, Ice Pond, had some of the highest nutrient concentrations which may be due to the lower mixing. Also, it is suspected that there are sufficient levels of ground water connection directly to the oceanic water which will reduce the biological effects of the ponds.

Because of the high recreational use of the estuarine ponds the concentrations of resident pathogenic microbes are of public concern. Hawaii Department of Health has been monitoring pathogenic densities in limited costal location on Hawaii Island but more regular and expansive monitoring should be conducted. Many of the CFU values for E. faecalis were well over the state standard of 7 CFU and some of the C. perfringens measurements were over the state standard of 5 CFU. The bacterial counts also showed high frequency and range of fluctuations in concentrations of C. perfringens and E. faecalis which demonstrate the need for more regular monitoring and locating the source of contamination. The use of two microbe indicators appears to be effective since each species are found in different environmental conditions. The high CFU values clearly demonstrate the poor water quality, possibly from sewage and runoff, of these recreational ponds.

To increase the overall power of this study, more nutrient and bacterial samples should be taken as well as more surveys of benthic organisms. Our results appear variable but this could be attributed to a variable collection plan. All parameters were measured monthly and not weekly. If the parameters were measured weekly a clearer pattern may exist. The bacteria samples were taken throughout the month and not during one week. By collecting the samples at different times of the month, other factors may have a larger role in the pattern of bacteria abundance. There was also high surge and heavy rains during the study. This could have altered the usual patterns of nutrients, bacteria and benthic organisms. When conducting a study this expansive and sampling intensive, there are many possible sources of error and many unaccounted for factors that could affect results.

Overall, the data acquired in this study seems to be highly variable, highlighting the need for more frequent sampling to monitor the water quality of the Keaukaha coast. However, our preliminary data show that the flora in these estuarine ponds do not appear to be remediating the input of terrigenous nutrients before freshwater mixes with oceanic water. The nutrient loads in rapidly transported precipitation and anthropogenic runoff are thus directly being exposed to the fringing coral reefs of Keaukaha coast. Although the sources of these nutrients are unclear at this point, the influence of fecal contamination is strongly indicated by the high levels of E. faecalis and C. perfringens. Continued monitoring of water quality in this populated coastal line is highly recommended for both public safety and management of coastal reefs.


Hsieh, Clement. (2000) Analysis of Nitrogen Sedimentation in the Advanced Integrated Wastewater Pond System (AIWPS) at Richmond Field Station, U.C. Berkeley. Jan. 26,

2006. (http://ist-socrates.berkeley.edu/~es196/projects/2000final/hsieh.pdf).

Natural Resources Defense Council (NRDC). (2004) Testing the Waters: A Guide to Water

Quality at Vacation Beaches. H1-H3.

Petit, Laetitia., Gibert, Maryse., and Michel R. Popoff. (1999) Clostridium perfringens:

toxinotype and genotype. Trends in Microbiology 7, 104-110.

Sasaki, Y., Yamamoto, K., Kojima, A., Norimatsu, M., and Y. Tamura. (2000) Rapid

Identification and differentiation of pathogenic clostridia in gas gangrene by polymerase

chain reaction based on the 16S-23S rDNA spacer region. Research in Veterinary Science 69, 289-294.

Shayler, Hannah A., Joel E. Creswell, Rebecca T. Johnson, Rebecca D. Karasack. (2000) The

Bioavailability of Dissolved Organic Nitrogen in Coastal Pond Waters. Jan. 26, 2006.


Shibata, Tomoyuki., Solo-Gabriele, H.M., Fleming, L.E., and Elmir, S. (2004) Monitoring

marine recreational water quality using multiple microbial indicators in an urban tropical environment. Water Research 38, 3119-3131.

United States Environmental Protection Agency (U.S. EPA). (2004) EPA Proposes More

Protective Water Quality Standards for Bacteria.