Hilo Bay Research Articles

Spatially distributed water quality responses to freshwater discharge in a tropical estuary, Hilo Bay, Hawai'i.

Hilo Bay, Hawai'i, is an estuary of great importance to its neighboring coastal community, but it is threatened by impaired water quality indicated by excessive turbidity and chlorophyll a associated with river discharges of sediments and nutrients. The Wailuku River in the western half of the bay is the primary source of freshwater discharge, hypothesized here to form a surface water-dominant half of the bay with different water quality traits than the groundwater-dominant, eastern half of the bay where the spring-fed Wailoa River discharges. The water quality of both halves of the bay over different flow conditions of the Wailuku River is examined in this study using spatially distributed water quality sampling which collects hundreds of samples in either half of the bay at a distance of about every 40m. The dense sample shows significant differences between the two halves of the bay, with greater salinity dilution and turbidity in the surface water-dominant area. Both salinity and turbidity have a predictable relation to discharge, with salinity decreasing and turbidity increasing in higher flow conditions. Chlorophyll a, however, has a more complex relation to discharge, as chlorophyll a concentrations are greatest in high-flow conditions, but this may be because the water quality samples were collected in different seasons. Furthermore, significantly greater chlorophyll a concentrations in the groundwater-dominant half of the bay in low-flow conditions show that discharge may be spuriously correlated to chlorophyll a, and further studies of the effects of surface water discharge on chlorophyll a concentrations are warranted.

A comparison of soil Staphylococcus aureus and fecal indicator bacteria concentrations across land uses in a Hawaiian watershed.

Staphylococcus aureus, methicillin-resistant S. aureus (MRSA), and fecal indicator bacteria (FIB; Enterococcus spp., Clostridium perfringens) concentrations increase in Hawaiian streams and estuaries following storms and pose a health threat to recreational water users. To reduce this risk, watershed bacteria sources need to be identified for management actions. This study's goals were to identify soil bacteria sources among different land uses and to determine if their concentrations were associated with different soil properties. Soil samples were collected three times on 24 d between October 2017 and November 2018 at urban, agriculture, and native-forest land uses in the Hilo Bay watershed, Hawai'i Island, Hawai'i. Soil bacteria concentrations were quantified using culturing techniques with selective media. Staphylococcus aureus, MRSA, and FIB were present in soil from all land uses. Bacteria concentrations were highest in urban soils and lowest in native-forest soils, with up to three orders of magnitude differences among land uses. Staphylococcus aureus, MRSA, and FIB soil concentrations were positively correlated with each other and with soil temperature and pH, but inversely correlated with soil moisture and organic matter content. Our results demonstrate that soils are a watershed bacteria source and that some soil properties affect their concentrations. Identifying these sources is critical for implementing management actions to reduce pathogen loads to estuaries and transmission to recreational water users.

Shallow coastal water turbidity monitoring using Planet Dove satellites

AbstractTurbidity monitoring in shallow coastal waters is fundamental to marine ecosystem research, management and protection. Satellite‐based water turbidity monitoring can be conducted at a greater spatial extent and higher temporal frequency than field measurements. The new Planet Dove satellite constellation has a daily revisit frequency and higher spatial resolution than Sentinel or Landsat satellites, allowing Planet Dove to track water turbidity dynamics in greater detail when suitable atmospheric correction is provided. We developed a new shallow coastal water turbidity estimation algorithm for Planet Dove and similar multi‐spectral satellites. Our algorithm accounts for bottom reflectance in total water‐leaving radiance to derive turbidity values in shallow coastal waters. We tested the algorithm with data from 235 Dove satellite images at five sites with different water conditions (Pelekane Bay, Big Island, Hawai‘i; Hilo Bay, Big Island, Hawai‘i; Kilo Nalu and Ala Wai, O‘ahu, Hawai‘i; Fagatele Bay, American Samoa; Vieques Island, Puerto Rico). We then validated satellite‐derived turbidity results (RMSE = 0.79–1.12 FNU [Formazin Nephelometric Unit]) using 75 days of field‐measured data, ranging in turbidity from 0.1 to 11.6 FNU in the five sites. Results show that our algorithm accurately detects turbidity in critical nearshore environments. In Hawai‘i, we used ~6700 Dove images to support a weekly turbidity monitoring study at a large geographic scale. We found this new, shallow‐water algorithm can be effectively applied to Dove satellite data to monitor water turbidity at high temporal resolution.

Diversity and health risk potentials of the Enterococcus population in tropical coastal water impacted by Hurricane Lane.

Hurricane-caused stormwater runoffs transport diverse terrestrial pollutants, adversely impact microbiological water quality, and introduce fecal and other pathogens to coastal water environments. This study investigated the genotypic diversity, phylogenetic composition, antibiotic resistance patterns, and virulence gene repertoire of the Enterococcus population in the Hilo Bay coastal water after the immediate impact of Hurricane Lane. DNA fingerprinting of Enterococcus isolates exhibited large genotypic diversity, while 16S rRNA gene sequencing identified four major species, including E. faecalis (34.7%), E. faecium (22.4%), E. hirae (22.4%), and E. durans (18.4%). Four common enterococcal virulence genes (cylA, esp, asa1, and gelE) were detected in the Enterococcus population, with significant portions of E. durans (33.3%), E. faecalis (41.2%), E. faecium (36.4%), and E. hirae (27.3%) isolates possessing two or more virulence genes. Considerable antibiotic resistance to rifampin, erythromycin, tetracycline, and nitrofurantoin was detected in the Enterococcus population, with one E. durans isolate showing vancomycin resistance. The results indicate considerable health implications associated with Enterococcus spp. in the hurricane-impacted tropical coastal water, illustrating the needs for more comprehensive understanding of the microbiological risks associated with storm-impacted coastal water.

Immediate Impact of Hurricane Lane on Microbiological Quality of Coastal Water in Hilo Bay, Hawaii.

Hurricanes and associated stormwater runoff events are expected to greatly impact coastal marine water quality, yet little is known about their immediate effects on microbiological quality of near-shore water. This study sampled Hilo Bay immediately after the impact of Hurricane Lane to understand the spatial and temporal variations of the abundance and diversity of fecal indicator enterococci, common fecal pathogens, and antibiotic resistance genes (ARGs). Water samples from seven sampling sites over 7 days were collected and analyzed, which showed that the overall microbiological water quality parameters [enterococci geometric mean (GM): 6-22 cfu/100 mL] fell within water quality standards and that the temporal dynamics indicated continuing water quality recovery. However, considerable spatial variation was observed, with the most contaminated site exhibiting impaired water quality (GM = 144 cfu/100 mL). The Enterococcus population also showed distinct genotypic composition at the most contaminated site. Although marker genes for typical fecal pathogens (invA for Salmonella, hipO for Campylobacter, mip for Legionella pneumophila, and eaeA for enteropathogenic Escherichia coli) were not detected, various ARGs (ermB, qurS, tetM, blaTEM, and sul1) and integron-associated integrase intI1 were detected at high levels. Understanding the temporal and spatial variation of microbiological water quality at fine granularity is important for balancing economic and recreational uses of coastal water and the protection of public health post the impact of major hurricane events.

Spatial analysis of water quality parameters in Hilo Bay, Hawai'i, using a combination of interpolated surfaces and hot spot analysis.

Hilo Bay estuary, located on the northeastern side of Hawai'i Island, experiences variability in water quality parameters due to its numerous water inputs. This estuary experiences influxes of water from three sources: groundwater to the east, marine water from the north, and surface water from the Wailuku River to the west. High rainfall and river flow impacts Hilo Bay's water quality including salinity, turbidity, and chlorophyll a concentration. Here, maps of Hilo Bay water quality were examined to assess spatial patterns of these important parameters. Exploring the patterns of these water quality parameters by creating inverse distance weighted (IDW) interpolation surfaces of survey points and clusters based on hot spot analyses during low- and high-flow conditions showed statistically significant differences in spatial water quality in Hilo Bay. Water quality maps after a storm show (1) an overall decrease in salinity, (2) a river plume from the Wailuku River associated with a turbidity hot spot, and (3) a chlorophyll a hot spot offset from the river plume in the center of the bay. Using spatial analysis to analyze water quality throughout the entirety of Hilo Bay before and after storm events can lead to a better understanding of how this ecosystem is affected during these types of events, and furthermore, adopting this method of sampling and analysis allows for a greater representation of water quality all over the bay and can improve the monitoring of water quality in this important ecosystem.

Rainfall and Streamflow Effects on Estuarine Staphylococcus aureus and Fecal Indicator Bacteria Concentrations

Pathogenic bacteria in nearshore waters are a public health threat, and many have watershed sources. Hence, understanding direct and indirect causes of bacterial loading can improve awareness and watershed management. Rainfall‐driven runoff influences river discharge, affecting pathogen transport to the ocean. This study assessed pathogen loading to nearshore waters under varying weather conditions within Hilo Bay, Hawaii, from 2014 to 2017. Staphylococcus aureus, methicillin‐resistant S. aureus (MRSA), and fecal indicator bacteria (FIB) were quantified in the bay, rivers, and potential watershed sources using culture‐based methods. Relationships between their concentrations with rainfall, river discharge, and water quality data were examined. Staphylococcus aureus, MRSA, and FIB were present within Hilo Bay and its rivers, as well as road runoff, sewage, and soils; MRSA was less prevalent. Staphylococcus aureus and FIB concentrations increased with rainfall and river discharge. Turbidity and salinity were the best water quality parameters for predicting bacteria concentrations, with positive and negative relationships, respectively. Our results suggest that more intense storms, especially after longer dry periods between events, will increase S. aureus and FIB loads to nearshore waters, as storms comprise >80% of annual river loads. Our models can be used to assess recreational water users’ health risks and predict future water quality conditions with changing rainfall patterns.Core Ideas Staphylococcus aureus and FIB were present in the bay, rivers, road runoff, sewage, and soils. MRSA was less prevalent in the environment than S. aureus and FIB. S. aureus and FIB bay concentrations increased with rainfall and river discharge. S. aureus and FIB bay concentrations varied with bay salinity and turbidity. Bay bacteria concentrations can be modeled from hydrologic and water quality data.

The life history of the invasive mullet, Osteomugil engeli (Bleeker, 1858) in Hawaiian estuaries

Although mullets (Mugilids) are important estuarine fishes across the world, the biology of many species of mullet have yet to be determined. This is of concern for mullets that are introduced species with unknown impacts on native ecosystems. Here we assess the life history of Osteomugil engeli, an introduced species in Hawai’i. Two populations (Hilo Bay estuary, Hawai’i Island and Maunalua Bay estuary, Oahu) were compared to determine population demographics, spawning season, and size and age at maturity. There were no differences observed between life history metrics between populations sampled. O. engeli had year round spawning with recruits and adults co-occurring in similar habitats throughout the year. There was no difference in size (L50) and age (A50) at sexual maturity for females and males (L50: χ2 = 0.39, p value =0.53; A50: χ2 = 0.29, p value = 0.59), with O. engeli reaching sexual maturity at 140.1 mm (CI: 139.5–141.6 mm) and 208 days (CI: 204.2–220.9 days). We observed a 4:1 female to male sex ratio and females were found to be on average 23 mm larger than males (t = 7.19, p value<0.01). O. engeli life history is compared to native species in Hawai’i and among other Mugilids that have been introduced elsewhere. Life history comparisons suggest that introduced mullet populations are better able to respond and adapt to changing habitat and environmental conditions than the native mullet.

Ensemble models with uncertainty analysis for multi-day ahead forecasting of chlorophyll a concentration in coastal waters

ABSTRACTIn this study, ensemble models using the Bates–Granger approach and least square method are developed to combine forecasts of multi-wavelet artificial neural network (ANN) models. Originally, this study is aimed to investigate the proposed models for forecasting of chlorophyll a concentration. However, the modeling procedure was repeated for water salinity forecasting to evaluate the generality of the approach. The ensemble models are employed for forecasting purposes in Hilo Bay, Hawaii. Moreover, the efficacy of the forecasting models for up to three days in advance is investigated. To predict chlorophyll a and salinity with different lead, the previous daily time series up to three lags are decomposed via different wavelet functions to be applied as input parameters of the models. Further, outputs of the different wavelet-ANN models are combined using the least square boosting ensemble and Bates–Granger techniques to achieve more accurate and more reliable forecasts. To examine the efficiency and reliability of the proposed models for different lead times, uncertainty analysis is conducted for the best single wavelet-ANN and ensemble models as well. The results indicate that accurate forecasts of water temperature and salinity up to three days ahead can be achieved using the ensemble models. Increasing the time horizon, the reliability and accuracy of the models decrease. Ensemble models are found to be superior to the best single models for both forecasting variables and for all the three lead times. The results of this study are promising with respect to multi-step forecasting of water quality parameters such as chlorophyll a and salinity, important indicators of ecosystem status in coastal and ocean regions.

Large amplification of the 1906 Colombia–Ecuador earthquake tsunami in Hilo Bay induced by bay-scale resonance

Large amplification of the 1906 Colombia–Ecuador earthquake tsunami in Hilo Bay induced by bay-scale resonance

Effect of river flow on the quality of estuarine and coastal waters using machine learning models

This study explores the river-flow-induced impacts on the performance of machine learning models applied for forecasting of water quality parameters in the coastal waters in Hilo Bay, Pacific Ocean. For this purpose, hourly recorded water quality parameters of salinity, temperature and turbidity as well as the flow data of the Wailuku River were used. Several machine learning models including artificial neural network, extreme learning machine and support vector regression have been employed to investigate the river-flow-induced impact on the water quality parameters from the current time up to 2 h ahead. Following the input structure of the machine learning models, two separate models based on including and excluding the river flow were developed for each variable to quantify the importance of the flow discharge on the accuracy of the forecasting models. The performance of different machine learning models was found to be close to each other and showing similar pattern considering accuracy and uncertainty of the forecasts. The results revealed that flow discharge influenced the water salinity and turbidity of the bay in which the models including the river flow as input variables had better performance compared with those excluding the flow time series. Among the water quality parameters investigated in this research, river flow made the most and least improvement on the efficiency of the models applied for forecasting of turbidity and water temperature, respectively. Overall, it was observed that water quality parameters can be properly forecasted up to several hours ahead providing a potentially valuable tool for environmental management and monitoring in coastal areas.

Spatial and temporal microbial pollution patterns in a tropical estuary during high and low river flow conditions

Spatial and temporal patterns of coastal microbial pollution are not well documented. Our study examined these patterns through measurements of fecal indicator bacteria (FIB), nutrients, and physiochemical parameters in Hilo Bay, Hawai'i, during high and low river flow. >40% of samples tested positive for the human-associated Bacteroides marker, with highest percentages near rivers. Other FIB were also higher near rivers, but only Clostridium perfringens concentrations were related to discharge. During storms, FIB concentrations were three times to an order of magnitude higher, and increased with decreasing salinity and water temperature, and increasing turbidity. These relationships and high spatial resolution data for these parameters were used to create Enterococcus spp. and C. perfringens maps that predicted exceedances with 64% and 95% accuracy, respectively. Mapping microbial pollution patterns and predicting exceedances is a valuable tool that can improve water quality monitoring and aid in visualizing FIB hotspots for management actions.

Mapping Ocean Currents Through Human Observations: Insights from Hilo Bay, Hawai'i

Complex systems, such as ocean currents, occur at multiple temporal and physical scales require simultaneous analysis across a range of geographic scales. Presently, there are few available nearshore current maps or models accessible to managers or the public in Hawai'i despite the fact that predicting nearshore currents and processes is important for understanding many other social-ecological interactions. Maps of coastal ocean currents are difficult to create because of constant change and the limited availability of nearshore data. Maps are symbols of our collective knowledge frameworks, representing various geographic areas and features that humans utilize. Our objectives were to understand human observations of nearshore ocean currents and the ability of ocean observers to communicate this knowledge. In Hilo Bay, Hawai'i, we asked 30 experienced ocean users, based on their natural observations, to create ocean current maps that share their knowledge of the seascape, and important processes that define each area. We then compared the scale of human observations of the seascapes with automated coastal observatories. We find that ocean observers were able to communicate their knowledge regarding ocean currents on maps at multiple spatial scales, and particularly commented spatially at a 1:5000 map scale. Understanding differences and similarities between the human observation scale and the in-situ mechanical observatories enable a more complete understanding of small-scale oceanic environments.

Doak Cox and the Woodcut

Doak Cox and the Woodcut Amanda Morris (bio) The thing about a mirror is that it shows you only an exact replica of what used to be. It is the fluid record of the death of the present played back at the speed of light. It is the image from a camera’s viewfinder that never becomes a snapshot. When we look to a mirror to see who we are, it shows us who we were just a moment ago. His name from childhood is Tokitaro. His name is Shunro, then it is Taito, then it is Iitsu. It is Gakyo Rojin Manji, which means “The Old Man Mad about Art.” He is an artist, and each time he makes a major shift in his artistic style, he changes his name. Now his name is Hokusai. Hokusai’s father made mirrors and taught him an appreciation for elegance. As a young boy, he watched as his father painted intricate designs around mirrors and then mimicked the craft. Seventy-year-old Hokusai dips his brush into a vial of black ink. His movements are controlled and deliberate. Beneath his bald and neatly creased forehead, Hokusai’s dark eyes squint severely at the page. He places the tip of his brush on the edge of the page and sweeps upward toward the center. He is tracing a curve. His brush jets back and forth in short, precise movements. His painted curls and dips are the grasping fingertips of a frothy wave. He has painted this wave before. This is the thirty-seventh and final draft of his Great Wave. Hokusai has been dead for ninety-seven years. In another forty-five a crater on Mercury will be given his name. It is 1946. On April 1 an earthquake off the coast of Alaska sparks a tsunami in the Pacific Ocean. The tsunami hurtles toward Hawaii more than three thousand miles from the quake’s epicenter. It reaches the Big Island in five hours and announces its arrival with a roar. It breaks over Pier 1 in Hilo Bay harbor, knocking people into the water. The first wave towers thirteen stories over the [End Page 169] Hawaiian coast. A trail of smaller waves bursts from the ocean in every direction, killing dozens in its path, which stretches from Alaska to South America. A person in Santa Cruz, California, drowns. Fishing boats sink off the coast of Chile. In Hawaii alone, 159 people were killed by the tsunami. Days later, geophysicist Doak Cox scratches neat equations on a pad in his office. A cigarette burns short in the ashtray on his desk. Cox scratches his head unconsciously as he considers data from the earthquake and subsequent tsunami. Like the long equations he writes to explain otherwise inexplicable phenomena, he speaks in long sentences, stringing out clause after clause, chasing the strain of his imagination through ebbs and flows, drawing breathless conclusions. The United States government asks Cox to join a scientist from the Scripps Institution of Oceanography and another from the University of Hawaii in the first post-tsunami survey ever conducted. Each time I visit the Art Institute of Chicago, I pause in front of Hokusai’s Great Wave. The crashing wave frames Mount Fuji, dwarfed in the background of the image. Mount Fuji’s snowy peak is as white and clean as the threatening crest of the wave on the Pacific Ocean. The outermost ringlet of foam points directly to the mountain from above, as though poised to devour it. Nearly lost in the image are the boats. The viewer is meant to look first at the giant cresting wave, and then to the mountain safely nestled in the distance. I must have stood in front of the print fifty times before I saw the fishing boats. The boats and their crews of twelve men dressed in the same blues and whites of the ocean bow their bald heads toward the water. They strain against their oars and paddle for shore or for calmer waters, whichever they find first. Their long boats are built for speed to transport their catch. Fresh bonito was a prized delicacy and promised a payday worth the...

Development of wavelet-ANN models to predict water quality parameters in Hilo Bay, Pacific Ocean

The main objective of this study is to apply artificial neural network (ANN) and wavelet-neural network (WNN) models for predicting a variety of ocean water quality parameters. In this regard, several water quality parameters in Hilo Bay, Pacific Ocean, are taken under consideration. Different combinations of water quality parameters are applied as input variables to predict daily values of salinity, temperature and DO as well as hourly values of DO. The results demonstrate that the WNN models are superior to the ANN models. Also, the hourly models developed for DO prediction outperform the daily models of DO. For the daily models, the most accurate model has R equal to 0.96, while for the hourly model it reaches up to 0.98. Overall, the results show the ability of the model to monitor the ocean parameters, in condition with missing data, or when regular measurement and monitoring are impossible.

A Comparison of Water Quality Between Low- and High-Flow River Conditions in a Tropical Estuary, Hilo Bay, Hawaii

Effects of storms on the water quality of Hilo Bay, Hawaii, were examined by sampling surface waters at 6 stations 10 times during low-flow and 18 times during high-flow (storms) river conditions. The direction of a storm’s impact on water quality parameters was consistent among storms and most stations; however, direction of the impact varied with the parameter. High river flow conditions increased concentrations of nitrate and decreased those of dissolved organic nitrogen (N); effects on ammonium and particulate N were station specific. Storms also increased dissolved organic and particulate carbon (C) concentrations. Dissolved phosphorus (P) concentrations were not affected by high river flow events. Dissolved organic forms dominated the N, C, and P pools under both low- and high-flow river conditions. Soil-derived particles and fecal indicator bacteria increased during storms, while chlorophyll a concentrations and bacterial cell abundances decreased. Our results suggest that an increase in storms with global warming could impact water quality of tropical estuaries.

Effects of hydrological forcing on the structure of a tropical estuarine food web

River flow can impact which sources of particulate organic matter (POM) fuel estuarine food webs. Here, we used stable carbon (C) and nitrogen (N) isotope analyses to compare how contributions of different POM sources (terrestrial, estuarine, and marine) to the diets of zooplankton and juvenile fishes differed between low and high river flow conditions, as well as spatially across a tropical estuary, Hilo Bay, Hawaii, USA. Diets of zooplankton and juvenile fishes were affected by river flow conditions, but the magnitude and the change in the basal resources depended on the location of the station in the estuary relative to the ocean and the river mouths. Consumers from the station most isolated from the ocean and with groundwater and overland flow inputs, utilized a combination of estuarine and terrestrial POM during both low and high river flow conditions and exhibited less variability in their basal resources than stations with direct ocean exchange. Consumers from stations in the Bay most affected by ocean exchange and river inputs utilized a combination of estuarine, terrestrial, and marine POM during low flow conditions, but shifted to marine and terrestrial POM during high river flow conditions. This shift to using terrestrial POM during high river flow conditions was substantial and up to 40% higher than values measured in other estuaries. Factors suspected to be affecting which POM source(s) consumers use in Hilo Bay are gross primary production, biological availability of exported terrestrial OM, and estuarine bacteria biomass, all of which are affected by river flow. Overall, our results suggest that Hilo Bay's food web and possibly those from other tropical estuaries are vulnerable to changes in hydrology, which may be further enhanced by global climate change.

Surface Water Metabolism Potential in a Tropical Estuary, Hilo Bay, Hawai’i, USA, During Storm and Non-storm Conditions

Surface water gross primary production potential (pGPP), respiration (RESP), metabolism potential (pMET), and CO2 fluxes in Hilo Bay, Hawai'i, USA, were examined along two river plumes during storm (high-flow) and non- storm (low-flow) conditions. Significant differences in pGPP, RESP, and pMET were found between low- and high-flow conditions, with lowest rates of all processes occurring during high-flow conditions. CO2 fluxes were influenced by metabolic processes at all but one site, with the bay's surface waters being autotrophic and a sink for atmospheric CO2 during low-flow conditions and less autotrophic and a source of atmospheric CO2 during high- flow conditions. Significant differences in pMET were found between the two river plumes during low-flow conditions at spatial scales of 1.5 km; however, no differences between river plumes were found during high- flow conditions. Our study suggests that an increase in storms associated with global climate change could impact surface water metabolic dynamics of tropical estuaries.

Sweet Memories: The Laupahoehoe Train Museum, Hawai'i

How does the Laupahoehoe Train Museum on the Big Island of Hawai'i preserve local cultural heritage? In the past, sugar cane dominated the landscape while trains moved harvested cane from fields to mills and transported sugar to ships in Hilo Bay. Both the trains and the cane have disappeared: the Hawai'i Consolidated Railway chose not to rebuild after the devastation of the 1946 tsunami and the sugar industry faced its ultimate demise in the 1990s. This museum is not a typical train museum with ambitions to amass a significant collection of railway cars and railway artifacts arranged and interpreted systematically. Rather, it is a small museum that portrays and documents the history and artifacts of plantation life, and initiates award-winning community outreach programs. In Sweet Memories, Mehos illuminates the interconnected histories of the sugar and rail industries on Hawai'i and how this museum strives to make visible the invisible past.

Carbon dioxide dynamics in rivers and coastal waters of the “big island” of Hawaii, USA, during baseline and heavy rain conditions

The distributions of the partial pressure of carbon dioxide (pCO2) and total alkalinity (TA) were examined for a 6-month period in the Wailuku and Wailoa rivers and coastal waters of Hilo Bay on the west coast of the Island of Hawaii, USA. Main results for the largest and turbulent Wailuku River show in the watershed an oversaturation in CO2 with respect to atmospheric equilibrium and a CO2 undersaturation in the estuary. In the Wailoa river-estuary system, extremely high pCO2 values ranging from 1500 to 10500 ppm were measured with significant shifts in pCO2 from drought to flood period. In the two rivers, water residence time, groundwater inputs and occasional flood events are the predominant drivers of the spatial and temporal patterns in the distribution of pCO2. In Hilo Bay, CO2 oversaturation dominates and the bay was a source of CO2 to the atmosphere during the study period. TA is conservative along the salinity gradient, indicating calcification in the bay is not a significant source of CO2 to the atmosphere.