Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
i
Identifying Hotspots of Sewage Pollution in Coastal Areas with Coral
Reefs
Presented to the Faculty of the
Tropical Conservation Biology and Environmental Science Graduate Program
University of Hawaiʻi at Hilo
In partial fulfillment of the requirement for the degree of
Master of Science
In Tropical Conservation Biology and Environmental Science
August 2016
By
Leilani M. Abaya
Approved by:
Tracy N. Wiegner
Karla J. McDermid
Jonathan D. Awaya
i
Acknowledgments
I express my sincere gratitude to my advisor, Tracy Wiegner, and committee
members, Jonathan Awaya and Karla McDermid. Their assistance and guidance over the
last two years have helped tremendously in my growth as a scientist. I also sincerely
thank all my family and friends who supported my academic pursuits over the last few
years. Without your continuing inspiration, love, and guidance I would not be the person
I am today. I also thank the funding agencies that made this project possible: National
Oceanic and Atmospheric Administration (NOAA NA14NOS4820087), Puakō
Community Association (PCA), Kamehameha Schools, Pacific Internship Program for
Exploring Sciences (PIPES), Center for Microbial Oceanography: Research and
Education (CMORE), Sigma Xi, STEM Honors Program, and the Marine Science
Department at the University of Hawaiʻi at Hilo. I would also like to thank the following
co-authors: Steven Colbert, Kaileʻa Carlson, Lindsey Kramer, and Jim Beets.
I am grateful to all the interns, fellows, volunteers, undergraduate students, and
graduate students who assisted in the different aspects of this project: Maile Aiwohi,
Evelyn Braun, Ricky Tabandera, Carrie Soo Hoo, Devon Aguiar, Jazmine Panelo, Cherie
Kauahi, Serina Kiʻili, Chelsie Wung, Bryan Tonga, and Louise Economy.
In addition to those mentioned above, I thank the University of Hawaiʻi at Hilo
Analytical Laboratory, Northern Arizona University Stable Isotope Laboratory, and Craig
Nelson who all had a part in the sample analysis. I am thankful for my collaborators: The
Nature Conservancy (TNC) and the Coral Reef Alliance (CRA) who made it possible to
share this work with community members, and assisted in fieldwork and support.
* This thesis is written as two separate manuscripts which will be submitted for
publication*
ii
TABLE OF CONTENTS
PAGE
Overview abstract………………………………………………………………………....1
Chapter 1:
A multi-indicator approach to identify hotspots of shoreline pollution in coral reefs
Abstract…………………………………………………………………………………....3
Introduction………………………………………………………………………………..4
Methods……………………………………………………………………………………7
Results……………………………………………………………………………………10
Discussion………………………………………………………………………………..12
Conclusion……………………………………………………………………………….17
Tables…………………………………………………………………………………….19
Figures…………………………………………………………………………………....23
Literature Cited…………………………………………………………………………..30
Chapter 2:
Spatial distribution of sewage pollution on a Hawaiian coral reef
Abstract…………………………………………………………………………………..37
Introduction………………………………………………………………………………38
Methods…………………………………………………………………………………..41
Results……………………………………………………………………………………44
Discussion………………………………………………………………………………..45
Conclusion……………………………………………………………………………….49
Figures……………………………………………………………………………………50
Tables………………………………………………………………………………….....56
Literature Cited…………………………………………………………………………..57
iii
LIST OF TABLES
Chapter 1 Tables PAGE
1. Average ± SE and range of sewage indicators...............................................................19
2. Average ± SE of δ15N - NO3-, NO3
-+NO2-, NH4
+, and PO43- of N sources...................20
3. Average ± SE and range of nutrient concentrations and salinity...................................21
4. Summary of sewage score indicators.............................................................................22
Chapter 2 Tables
1. Average ± SE and range of nutrient concentrations and salinity...................................56
iv
LIST OF FIGURES
Chapter 1 Figures PAGE
1. Sewage indicator stations, dye tracer sites, and N sources in Puakō, Hawaiʻi………..23
2. Sewage indicator data from shoreline stations………………………………………...24
3. Summary of δ15N in macroalgal tissue relative to N sources…….…………………...25
4. Mixing plots of salinity and nutrients………………………………………….……...26
5. Correlation between C. perfringens and NH4+…………………………………….….27
6. Correlations between δ15N in macroalgal tissue and nutrients…….………………….28
7. Puakō sewage pollution score map……...…………..………………………………...29
Chapter 2 Figures
1. Offshore sampling stations in Puakō, Hawaiʻi………………………………………..50
2. Purging experiment results……………………………………………………………51
3. Sewage indicator data from surface and benthic waters ………………………...........52
4. Summary of δ15N in macroalgal tissue relative to N sources..….…………….………53
5. Pre- and post-deployment δ15N values in Ulva fasciata………….……………..…….54
6. Puakō sewage pollution score map……………………………………………………55
1
Overview Abstract
Sewage pollution threatens human and coral reef health. Study goals were to
identify sewage pollution hotspots through dye tracer tests, measurements of sewage
indicators, and development of a sewage pollution score, along Puakō’s (Hawaiʻi) reef.
Sewage was localized within 10 m of the shoreline and reached it within 9 hours to 3
days. Shoreline nutrient concentrations were two times higher than upland groundwater.
Sewage indicators were higher and more variable along the shoreline than on the reef,
and often greater than water quality standards. Shoreline δ15N macroalgal values were
indicative of sewage, while offshore values were indicative of soil or groundwater nitrate.
A sewage pollution score was created using several indicators that accurately identified
sewage pollution hotspots, as three dye tracer locations had the highest scores. Results
highlight the need for a multi-indicator approach and scoring system for identifying
sewage pollution hotspots to improve water quality.
2
A multi-indicator approach to identify hotspots of shoreline sewage
pollution in coral reef areas.
Chapter 1
3
Abstract
Globally, coral reefs are declining because of various stresses, with sewage
pollution being one of the most concerning. This study used a combination of sewage
indicators to investigate the presence of sewage pollution along the Puakō coastline in
Hawaiʻi. Fecal indicator bacteria (FIB) counts and nutrient concentrations were high and
variable along the Puakō shoreline, and stable nitrogen isotope (δ 15N) levels in
macroalgal tissue were within the sewage range. These indicators together with results
from dye tracer tests, documented sewage pollution in Puakō’s coastal waters. However,
sewage indicator data were not always in agreement in terms of the intensity and location
of sewage pollution. A sewage pollution score was created using a combination of
sewage indicators and nutrients to identify pollution hotspots. The scoring system
allowed locations of sewage pollution to be visualized in a holistic manner. The approach
used here will be useful to other coastal communities around the world who are
challenged with documenting sewage pollution in their nearshore waters.
4
Introduction
With more than 50% of the world’s population living near coastal areas, sewage
pollution has become a growing global concern. Sewage is introduced into coastal
environments from cesspools, septic tanks, injections wells, and sewage treatment plants
(Peterson & Oberdorfer 1985; Hunt 2006; Knee et al. 2008; Zimmerman 2010). Sewage
pollution can result in elevated levels of pathogens, hydrocarbons, nutrients, toxins, and
endocrine disruptors in nearshore waters (Wear & Thurber 2015), which can have
devastating consequences on the health of recreational water users. For example, human
exposure to sewage can to result in skin and urinary tract infections, hepatitis, and
gastroenteritis illnesses (Pinto 1999). In fact, globally, over 120 million gastroenteric
illnesses annually are associated with sewage contaminated waters (Shuval 2003). In
addition, sewage pollution can have detrimental effects on coastal ecosystems, especially
coral reefs. Coral reefs are one of the most valuable and biologically diverse ecosystems
in the world, contributing billions of dollars in ecosystem services worldwide (Wear &
Vega Thurber 2015). However, coral reefs are steadily declining from multiple stresses
including, one of the most harmful, sewage pollution (Wear & Vega Thurber 2015).
Sewage pollution contributes to coral disease (Vega Thurber et al. 2014; Yoshioka et al.
2016). For example, Serratia marcescens, a human pathogen in sewage, caused ~88%
decline in coral cover in the Caribbean (Sutherland & Ritchie 2004). Nutrient enrichment
from sewage alters coral growth rates, species distribution and abundance, diversity, and
prevalence and severity of disease (Pastorok & Bilyard 1985; Dollar & Grigg 2004;
Parsons et al. 2008; Vega et al. 2014). These nutrients also stimulate benthic algal
growth, which can result in a benthic phase shift from coral to macroalgal reefs (Hunter
& Evans 1995; Lapointe et al. 2005; Paytan et al. 2006).
As the human population continues to grow along the coast, monitoring water
quality for sewage pollution is essential. Measurements of fecal indicator bacteria (FIB)
are a widely used method to detect sewage. FIB are used to evaluate potential risk to
human health. Enterococcus is monitored in marine recreational waters by the United
States Environmental Protection Agency (USEPA) (USEPA 2003), and in the tropics, a
secondary indicator, Clostridium perfringens, is used (Davies et al. 1995; Fung et al.
5
2007). Clostridium perfringens is an anaerobic, spore-forming bacterium that does not
multiply in coastal waters, nor grow in tropical soils like Enterococcus (Hardina &
Fujioka 1991; Fung et al. 2007). Measurements of stable nitrogen (N) isotopic
composition (δ15N) in macroalgal tissues is another method to detect sewage pollution in
coastal waters (Umezawa et al. 2002; Savage 2005; Hsing-Jun Lin et al. 2007; Dailer et
al. 2012; Wiegner et al. 2016). Macroalgae only minimally discriminate between 14N and
15N. Therefore, they have similar isotopic compositions relative to their N sources
(Savage 2005). Sewage, in particular, has a very distinct stable N isotopic composition
compared to other sources, i.e, fertilizers, soils, groundwater, and ocean water (reviwed
in Wiegner et al 2016). Additionally, the macroalgal species often used in these studies as
bioindicators are opportunistic ones whose growth is stimulated by increases in N (Littler
& Littler 1980; Dailer 2010; Dailer 2012).
Measuring a single sewage indicator to assess sewage pollution in a location
could be misleading due to the variability associated with the different indicators. For
example, one study found that when using Enterococcus, the authorities would be more
likely to post beach advisories than if using C. perfringens (Shibata et al. 2004). In
addition, δ15N in macroalgal tissues can be highly variable due to inputs of N from many
sources with different N isotope compositions (Ochua-Izaquirre and Soto-Jimenez 2015).
Hence, it is imperative to measure multiple indicators of sewage to determine spatial and
temporal patterns of pollution with greater confidence. Few studies have done this to
date, and at most, only two indicators have been measured simultaneously.
Hawaiʻi is an ideal location to develop a multiple indicator sewage detection
approach. Hawaiʻi’s coastal waters and coral reefs have been plagued by sewage
pollution for decades, initially from outfalls in enclosed bays and today from more
diffuse and prevalent sources – cesspools (Pastorok & Bilyard 1985). Cesspools are a
major source of sewage pollution especially in rural areas, which comprise most of the
state. Hawaiʻi uses cesspools more widely than any other state (USEPA 2013), and has
only recently banned the installment of new cesspools. Over 110,000 individual
wastewater systems exist in the state, with the Hawaiʻi Island having nearly 49,000
(Whittier & El-Kadi 2014). A high-risk area where wastewater systems are likely
6
impacting nearshore waters on the Hawaiʻi Island is Puakō (Whittier & El-Kadi 2014).
Puakō is a coastal community that is home to some of the richest, most diverse reefs in
the state (Hayes et al. 1982). However, coral coverage has decreased from 80% in 1975
to 33% in 2010 at Puakō (Minton et al. 2012), with concurrent decreases in fish
abundance (49% - 69%), and increases in turf and macroalgal cover (38%) (HDAR
2012). Declining coral health and increases in disease prevalence and severity have also
been documented (Couch et al. 2014). While sewage pollution is thought to be one of the
culprits contributing to these ecosystem changes, the link between these conditions and
the presence of presence has not been made.
The goal of this study was to develop a multiple sewage indicator approach to
more accurately detect the presence of sewage in nearshore waters. To do this, four
indicators were measured along the Puakō coastline: 1) FIB, 2) nutrients (up mountain
and near the ocean), 3) δ15N in macroalgal tissue, and 4) fluorescein dye tracer releases
from cesspools. Dye tracer studies have been used to effectively illustrate the connection
between hydrogeologic features (Gaspar 1987). Based on the concentration, dye tracer
studies can be used to calculate transit time, flow rates, and dilution in the aquifer during
transport between two points. Fluorescein, a non-toxic organic dye, has strong
fluorescence, which can be used as a tracer with detection levels as low as 1 ppb (Gaspar
1987; Reich et al 2001). In addition, to identify sewage pollution hotspots along the
coast besides homes where dye tracer studies were conducted, a sewage pollution score
was developed.
This is the first study to use a combination of techniques and indicators to
examine shoreline sewage pollution, as most studies to date have used a single one, or
two at most, to assess sewage pollution. Increasing human population in coastal areas is
causing sewage to be a worldwide water quality issue (Wear & Thurber 2015).
Developing an approach to assess sewage pollution is imperative for improving human
and nearshore ecosystem health. The multiple sewage indicator approach and sewage
pollution score developed through this research will be an invaluable management tool
for coastal areas. Information gathered through this approach will allow managers to
7
identify hotspots of sewage pollution and develop effective wastewater treatment
solutions for their coastal communities.
Methods
Site Description
This study was conducted along the Puakō coast (Fig. 1), which is primarily
comprised of basalt from the flow of Mauna Loa Volcano (Kaniku Flow in 1859) and is
mostly aʻa with some pahoehoe. The annual rainfall in this region ranges from 250 – 750
mm. Infiltration of rainwater into the aquifer is high due to the permeable substratum, and
average submarine groundwater discharge (SGD) ranges from 2083 - 2730 L m-1 h-1
(Paytan et al. 2006).
Puakō is a residential community consisting of 207 lots, of which 163 have
homes, along a 3.5 km stretch of the south Kohala coastline. The population is growing at
a rate of 6.9% per year (Minton et al. 2012). At Puakō, 47 homes have cesspools and 139
have conventional septic tanks with leach fields (Schott 2010). The entire coastline is
accessible to the public and is frequently used for recreational activities such as fishing,
surfing, diving, and snorkeling. Presently, there is one development up-slope of Puakō,
Waikoloa Village, with over 4,800 people, whose homes have onsite sewage disposal
systems (OSDS) (U.S. Census Bureau 2000).
Station Selection
To select shoreline stations for sampling, a salinity survey was conducted using a
YSI 6600 V2 multi-parameter sonde interfaced with a Garmin etrex Global Positioning
System (GPS). This was completed in the summer of 2014 during low tides to capture
maximum groundwater input. From this survey, sixteen shoreline stations were chosen
with varying salinity (Fig. 1).
FIB & Nutrient Analyses
Water samples were collected four times at all 16 stations and analyzed for FIB,
nutrients, and salinity. Samples were taken between November 2014 and July 2015
during low tide conditions near sunrise as sunlight has been shown to affect the survival
of bacteria (bactericidal effect) (Fujioka et al. 1981). FIB Enterococcus was analyzed
using the Enterolert MPN method (IDEXX Laboratories Inc) and C. perfringens was
8
quantified using a membrane filtration technique (Bisson & Cabelli 1979). Water samples
were filtered through pre-combusted (500°C for 6 hours), 0.7-µm pore size GF/F filters
(WhatmanTM), and stored frozen until analysis at University of Hawaiʻi at Hilo Analytical
Laboratory. Nutrients were analyzed on a Pulse TechniconTM II autoanalyzer using
standard methods (NO3- + NO2
- [Detection Limit (DL) 0.07 µmol L-1, USEPA 353.4],
NH4+ [Detection Limit (DL) 0.36 µmol L-1, USGS I-2525], PO4
3- [DL 0.03 μmol L-1,
Technicon Industrial Method 155-71 W], total dissolved phosphorous (TDP) [DL 0.5
μmol L-1, USGS I-4650-03], and H4SiO4 [DL 1 μmol L-1, USEPA 366]) and reference
materials (NIST; HACH 307-49, 153-49, 14242-32, 194-49). Total dissolved nitrogen
(TDN) was analyzed by high-temperature combustion, followed by chemiluminescent
detection of nitric oxide (DL 5 μmol L-1, ASTM D5176, Shimadzu TOC-V, TNM-1)
(Sharp et al. 2002). Salinity was measured at the time of water collection using a YSI Pro
2030 multi-parameter probe.
δ15N Analyses
At the time of water sample collection, the most abundant macroalgae were
collected at all stations and analyzed for δ15N (Fig. 1). Macroalgal tissues were placed on
ice during transport to the laboratory, where tissues were rinsed with deionized water.
Subsamples of macroalgae were preserved as voucher specimens and identified using an
OlympusTM CH30 microscope and identification books (Abbott 1999; Abbott & Huisman
2004). The remainder of the samples were dried at 60° C until a constant weight was
achieved, ground and homogenized using a Wig-L-Bug grinding mill, and ~2 mg of the
macroalgal tissue were folded in 4x6 mm tin capsules for stable isotope analysis.
Macroalgal tissues were analyzed for δ15N using a Thermo-FinniganTM Delta V
Advantage isotope ratio mass spectrometer (IRMS) with a Conflo III interface and a
CostechTM ECS 4010 Elemental Analyzer located at the University of Hawaiʻi at Hilo
Analytical Laboratory. Data were normalized to USGS standard NIST 1547. Isotopic
signatures were expressed as standard (δ) values, in units of parts per mil (‰), and
calculated as [(Rsample – Rstandard) / Rstandard] x 1000, where R = 15N/14N.
To determine the sources of N being used by the macroalgae, δ15N - NO3- of
potential sources were measured. Nutrient concentrations (NO3- + NO2
-, NH4+, and PO4
3-)
9
in these sources were also measured. Sources sampled included cesspools (n = 3), high
elevation groundwater wells (n = 3), low elevation groundwater wells (n =7), ambient
seawater (n = 2), and soil (n = 3) from under Kiawe trees (Prosopis pallida). Kiawe, an
introduced N2-fixing tree, found widely on the leeward coasts on the Hawaiʻi Island, and
contributes N to soil and groundwater (Gallaher & Merlin 2010; Dudley et al. 2014). Soil
was collected directly under the Kiawe trees, dried, and then shaken overnight with
reagent-grade water. N source water samples were collected at several locations to assess
spatial variability. All N source samples were filtered through a 0.22-μm cellulose acetate
filter (WhatmanTM) and frozen until analysis. δ15N - NO3- samples were analyzed on a
Thermo-FinniganTM Delta Plus IRMS with data normalized to U.S. Geological Survey
(USGS) standards (USGS32, USGS34, USGS53) at Northern Arizona University Stable
Isotope Laboratory. IAEA-NO3 was used as a check standard. Fertilizer values were also
used in this study that was collected from a previous sewage study on the Hawaiʻi Island
(Wiegner et al 2016). To determine their N sources, the δ15N macroalgal tissue values
were plotted relative to δ15N source values (Derse et al. 2007; Wiegner et al. 2016).
Dye Tracer Tests
Dye tracer tests were conducted to determine the hydraulic connectivity between
the OSDS at four oceanfront homes. Tests were conducted along the southern portion of
Puakō’s coastline as nearshore waters here were relatively fresh. Three homes had
cesspools, and one had a fractured aerobic treatment unit tank. Two homes were occupied
during the tests and the other two were not. At each home, the closest point where dye
could be delivered to the OSDS was identified, and 500 - 1000 g of high purity
fluorescein dye (Amresco Fluorescein Sodium Salt) were injected over approximately 10
h. Each hour, 50 or 100 g of dye were mixed with 20 L of tap water and slowly added to
the OSDS. Additional tap water was added throughout the day and its volume recorded
to calculate an initial dye concentration.
Five to six stations were identified in front of each home and adjacent properties,
representing three to four springs of varying salinity and two stations with higher salinity
and no apparent freshwater input. Samples for fluorescence were collected at each station
before and during dye tracer tests in an opaque brown high density polyethylene (HPDE)
10
bottle, pre-rinsed with sample water, to prevent photodegradation, and stored at 4°C until
analysis. During the first 12 h of the dye test, samples were collected every two h to
identify any fast-flow pathways. Afterwards, two samples were collected at each station
within an hour of the lowest-low water each day for up to 14 days.
To quantify the concentration of fluorescein, samples were brought to room
temperature, filtered (Whatman GF/F), and analyzed using a Turner AU10 fluorometer in
the dark. The detection limit for our analysis was 0.95 ppb (USEPA 40 CFR 2011).
When salinity was not measured in the field, conductivity of samples was measured in
the laboratory (Orion Star) and converted to PSS-78 salinity (Unesco 1981).
Data Analyses
To determine if FIB, δ15N values, and nutrients in macroalgal tissue differed
among stations, a one-way analysis of variance (ANOVA) was used. Correlations were
used to evaluate associations between FIB, δ15N values, nutrients, and other water quality
parameters. Data were tested for normality and equal variances. If assumptions were not
met for parametric analyses, log transformations were used. Statistical analyses were
conducted using Minitab17 (2010) with an α=0.05.
Mixing plots of nutrient concentrations and salinity were also used to determine
nutrient sources (freshwater vs. ocean) to coastal water bodies, and to examine their
behavior as freshwaters and ocean waters mix. Nutrient concentration data on mixing
plots were compared to a theoretical mixing line connecting the freshwater and ocean end
members in that system. When nutrient concentration data fell on the mixing line, the
nutrient was described as behaving conservatively – that is dilution was the only thing
affecting the nutrient concentration in the nearshore waters. When data fell above or
below the mixing line, the nutrient was described as behaving non-conservatively, with
some source adding the nutrient to the water or some process removing it during mixing.
Results
Sewage Indicators
Contrasting patterns were seen among sewage indicators along shoreline stations.
Enterococcus ranged from 18 to 2777 MPN/100 mL and counts significantly differed
among stations (p =0.04), with station 13 (Average ± SE) (2777 ± 1806 MPN/100 mL)
11
having the highest counts (Fig. 2, Table 1). Clostridium perfringens values ranged from 2
to 12 CFU/100 mL and counts were similar among stations, averaging 5 ± 3 CFU/100
mL across all stations (p =0.06) (Fig. 2, Table 1). The most prevalent macroalgal species
along the shoreline were Ulva fasciata, Cladophora spp., and Gelidiella acerosa. The δ
15N in macroalgal tissue ranged from 4.23‰ to 11.88‰ across all 16 shoreline stations,
and δ 15N significantly differed among stations (p <0.0001) (Fig. 2,Table 1), with stations
3 (10.55‰ ± 0.15) and 4 (11.88‰ ± 0.32) being the most enriched (Fig. 2). Six out of 16
stations fell within the sewage δ 15N - NO3- range (Table 2), including stations 3, 4, as
well as 5 (7.21‰ ± 2.01), 6 (7.77‰ ± 1.29), 7 (8.35‰ ± 0.73), and 13 (7.74‰ ± 1.93)
(Fig. 3). The remaining stations fell within the high and low elevation groundwater
ranges (Fig. 3). NO3- + NO2
- concentrations were lower in high elevation wells (93.9
µmol/L ± 4.4) compared to the lower elevation wells (130.1 µmol/L ± 6.7) (Table 2). In
addition, PO43- and NH4
+ concentrations were similar between high (2.5 µmol/L ± 0.2; 5
µmol/L ± 1, respectively) and low elevation wells (2.5 µmol/L ± 0.5; 5 µmol/L ± 1,
respectively) (Table 2). NO3- + NO2
-, TDN, PO43-, TDP, and H4SiO4 concentrations
significantly differed among shoreline stations (p <0.001) (Table 3). Station 4 had the
highest concentrations, and station 15 had the lowest concentrations for all nutrients,
except H4SiO4. H4SiO4 had the highest concentrations at station 14 (652 µmol/L ± 174),
while station 6 (95 µmol/L ± 43) had the lowest (Table 3). NH4+ concentrations were
similar across all shoreline stations (p >0.06). Salinity also varied across stations (p
<0.01), with stations 2 (7.12 ± 0.61) and 14 (6.43 ± 0.63) being the freshest (Table 3).
Nutrient concentrations (NO3- + NO2
-, TDN, PO43-, TDP, and H4SiO4,) were also
inversely correlated with salinity (p <0.01). Mixing plot analysis revealed conservative
mixing of groundwater-derived nutrients (NO3- + NO2
-, TDN, PO43-, TDP, and H4SiO4)
with seawater, except for a few stations that consistently fell well above the theoretical
mixing line (Fig. 4). These included stations 3, 4, and 7. NH4+ displayed non-
conservative mixing (Fig. 4).
Correlations Among Sewage Indicators
Most sewage indicators were not correlated with each other; however, δ15N in
macroalgal tissue values were correlated with C. perfringens and nutrient concentrations.
12
Specifically, C. perfringens was positively correlated with NH4+ (p =0.02) (Fig. 5) and δ
15N in macroalgal tissue was positively correlated with NO3- + NO2
- (p <0.001), TDN (p
<0.001), and PO43- concentrations (p <0.001) (Fig. 6).
Dye Tracer Tests
Dye was visually observed at the shoreline at all four sites. For each test, there
was only one spring with dye, which was located on the beach in front of the property.
The groundwater discharge at these springs dispersed over an area between 0.25 and 4
m2. Initial breakthrough of dye at the shoreline, calculated based on the start of the dye
tracer test until the first appearance of dye at the shoreline, ranged from 9 h to 3 d. Three
of the homes had comparable flow rates between 4 and 14 m/day; the OSDS at one home
was remarkably faster, where dye in groundwater traveled 76 m/day. Based on the
dilution of the dye, the maximum fraction of sewage in the freshwater at the shoreline
varied from <0.02% to 0.14%, depending on how much mixing occurred before
discharge at the shoreline.
Discussion
Sewage Indicators
FIB are used by federal and state regulatory agencies to monitor impaired
recreational waters. At Puakō, Enterococcus counts differed among shoreline stations,
with station 13 having the highest counts. Additionally, when comparing our average
values of Enterococcus counts to Hawaiʻi Department of Health’s (HDOH) single sample
maximum of 104 CFU/100 mL, 13 out of 16 stations exceeded this threshold (Fig 2).
However, Enterococcus has been shown to vary spatially, temporally, seasonally, and to
be tidally influenced (Shibata et al. 2004; Maiga et al. 2009; Shibata et al. 2010; Nnane et
al. 2011; Converse et al. 2012). In addition, Enterococcus has been found to persist in
tropical soils (Hardina & Fujioka 1991; Byappanahalli & Fujioka 1998; Byappanahalli &
Fujioka 2004), and thus, may not be indicative of sewage, but possibly of a soil source.
However, soils are an unlikely source of Enterococcus at Puakō as the area generally
lacks soil and the substratum is primarily basalt. This, in combination with the above
factors, indicate that Enterococcus may not be the best indicator because of its high
variability. However, the levels that were found along the shoreline were extremely high
13
compared to state standards and may be indicative of sewage. While C. perfringens
counts did not vary among stations, 11 of the 16 stations fell above the recommended
standard to HDOH for marine recreational waters (5 CFU/100 mL) (Fujioka et al. 1997).
Additionally, using the Fung/Fujioka C. perfringens scale for sewage pollution based on
single sample maximums (Fung et al. 2007), only four of our stations (stations 7, 11, 14,
and 15) were indicative of non-point sewage contamination (>10 CFU/100 mL). The
remaining five stations fell below this range, and are classified as not being polluted by
sewage. While only certain stations had C. perfringens counts within the range for non-
point sewage pollution, the correlation between C. perfringens and NH4+ suggest that the
sewage pollution may be more pervasive, as these two parameters are associated in
anaerobic conditions, which are often found in OSDS.
While FIB are used to detect sewage, their application is primarily for assessing
human health hazards for recreational water users. δ 15N values in macroalgal tissue, on
the other hand, are used to determine N sources to coastal waters including sewage
(Costanzo et al. 2005; Lapointe et al. 2005; Savage 2005; Derse et al. 2007; Dailer et al.
2012). Typical sewage values range from +5‰ to +20‰ (Kendall 1998; Hunt 2006;
Dailer et al. 2010; Derse et al. 2007; Xue et al. 2009), and values from the cesspools in
our study fell within this range (Table 3). δ 15N in macroalgal tissue along the Puakō
shoreline ranged from 4.23‰ to 11.88‰, with six out of 16 shoreline stations falling
within the range for sewage (Table 4, encompassing SE of averages). Stations 3 and 4
had the most enriched δ 15N macroalgal tissues, highlighting two potential sewage
pollution hotspots. However, past studies have found that macroalgae assimilate N more
rapidly under low NO3- concentrations (Fujita 1985), and that δ 15N in macroalgal tissue
can be underestimated by up to 6‰ in waters with high NO3- concentrations (>10
μmol/L) (Swart et al. 2014). All of the stations had NO3- + NO2
- concentrations
exceeding 10 µmol/L, suggesting that the δ 15N macroalgal values may be
underestimated. If this is the case, then all 16 stations fall within the sewage range. In
contrast, other studies have found that tissue from opportunistic macroalgal reflects the
nutrient concentrations in the water column (Fong et al. 1994). Two out of the three taxa
(Ulva fasciata and Cladophora spp.) collected along the Puakō shoreline were
14
opportunistic macroalgae; however, during sample processing, tissue from all three
macroalgae species were combined for δ 15N analysis, and G. acerosa is not considered
an opportunistic species. Additionally, NO3- + NO2
-, TDP, PO43-, and TDN
concentrations were nearly 8x greater at station 4 compared to all stations, this pattern
was also seen with δ 15N in macroalgal tissue. These patterns further suggest that station
4 is a hotspot of non-point sewage pollution at Puakō. Correlations between δ 15N in
macroalgae and nutrient concentrations also suggest that some portion of the nutrients’
concentrations are derived from sewage (Fig. 6).
Hydrology
At Puakō, a large portion of the nutrient concentration data for NO3-+NO2
-, TDN,
PO43-, and TDP fell on the theoretical mixing line, with highest concentrations at the
lowest salinities. This result suggests that high elevational groundwater is a source of
nutrients to Puakō’s coastal waters and that they are behaving conservatively as
groundwater and ocean water mix at the shoreline. This pattern, in part, explains the lack
of associations between sewage indicators and salinity, as a large portion of groundwater
nutrients discharging at the shoreline is from a nutrient source other than sewage. The
conservative mixing nutrient patterns observed at Puakō have been documented
elsewhere on Hawai‘i Island in coastal areas with groundwater inputs (Knee et al. 2008).
In contrast to the majority of our shoreline stations, data for stations 3, 4, and 7
consistently fell above the theoretical mixing line, suggesting there is a localized source
of nutrients in those areas. The likely source is the OSDS as our dye tracer tests
demonstrated that OSDS at these stations were leaking, and that the travel time from the
homes to the shoreline was 9 h to 3 d. Additionally, dye was only observed seeping out
during low tide and was localized within 10 m of the shoreline. δ15N-NO3- at these
shoreline stations clearly fell within our measured δ15N-NO3- sewage range at Puakō
(Table 2), as did the δ15N in the macroalgal tissue (Fig. 3). These results provide insight
to the hydrology and geology at Puakō, where the fracture system within the basalt
determines the flow path of the sewage from the OSDS to the shoreline, and affects the
time of travel. Two other factors affecting sewage inputs are weather and house
occupancy. On the only rainy sampling day during this study (March 4, 2015), all three
15
stations (3, 4, and 7) fell above the mixing line. This result illustrates that precipitation
inputs enhanced the connection between the OSDS with the shoreline seeps through
increased groundwater discharge. In contrast, there were some sampling dates on which
nutrient concentrations for stations 3 and 7 fell on the mixing line. We suspect that on
these dates, homes at these stations were not occupied, and therefore, there the OSDS
were not being used.
The δ15N-NO3- and NO3
- concentration data from the groundwater wells and
shoreline stations also provided another insight into the hydrology of the Puakō
watershed and coastal nutrient sources. The δ15N-NO3- became increasing enriched in
15N moving downslope to the Puakō shoreline. The change in the δ15N-NO3- from the
high to low elevational groundwater wells suggests a change in NO3- source from forest
soil to sewage. It is possible that sewage is contaminating the lower elevational
groundwater as an upslope development (Waikoloa Village) has over 4,800 people whose
homes have OSDS (U.S. Census Bureau 2000). Additionally, NO3- concentrations
increase ~40 µmol/L from the high to low elevational groundwater wells (Table 3).
Lastly, the 15N enrichment in the δ15N-NO3- from the lower elevational groundwater
wells to the shoreline seeps suggests that additional sewage from Puakō homes is
contaminating the groundwater before it is discharged along the shoreline. To understand
the relative percent contributions of these two different communities to sewage pollution
along Puakō’s shoreline, more δ15N-NO3- data and a mixing model capable of
determining source contributions is needed (Wiegner et al. 2016). With this additional
information, informed decisions about management actions can be made.
Development of a Sewage Pollution Score
As this study and others have shown, sewage indicators can provide conflicting
information on the intensity and location of sewage pollution along the shoreline. In this
study, Enterococcus counts were highly variable among stations, with some exceeding
HDOH standards, and station 13 having the highest counts. In contrast, C. perfringens
counts were similar among stations, but averages for stations 7, 11, 14, and 15 were in the
non-point source sewage pollution range (Fung et al. 2007). Additionally, δ 15N in
macroalgal tissue were found to be highly variable along the shoreline, with six stations
16
falling within the range of our sewage source value (Fig. 2, Table 3). Previous studies
have confronted similar issues with their sewage indicator data (Shibata et al. 2004;
Yoshioka et al. 2016). Hence, creating a sewage pollution score using sewage indicators
may be a way to more holistically assess sewage pollution in coastal waters. Water
quality scores and indices have been used successfully in the past to assess healthy water
quality conditions for both humans and ecosystems (Zambrano et al. 2009; Wang et al
2015). One successful study was able to examine four different regions utilizing
physiochemical and nutrient parameters to relate human activities to water quality,
highlighting areas in need for better wastewater management.
To better assess sewage pollution conditions along the Puakō shoreline, a scoring
system was developed using sewage indicators (FIB, δ15N macroalgae, and nutrients).
The scoring system had three levels for each indicator: level 1 = low, level 2 = medium,
and level 3 = high. Levels for each indicator were based on established standards or
literature information (Table 4). Specifically, the scoring system used HDOH’s single
sample maximum for Enterococcus counts in marine waters (HDOH 2014), the
Fung/Fujioka C. perfringens scale for sewage pollution (Fung et al. 2007), δ 15N values
in macroalgal tissue for different N sources (reviewed in Wiegner et al. 2016), and
HDOH’s water quality standards for nutrient concentrations in open coastal waters (NO3-
+ NO2-, NH4
+, TDP) (HDOH 2014) (Table 4). Nutrient concentration standards for the
wet criteria were used because the freshwater inputs along the Puakō shoreline ranged
from 2083-2730 L m-1 h-1 (Paytan et al. 2006), an order of magnitude larger than the
baseline for the wet criteria (>294 L m-1 h-1). Two dissolved inorganic forms of N were
chosen for the score system rather than TDN because it contains DON and there are no
well-established sewage pollution patterns with this constituent. TDP was used as the
phosphorous water quality indicator since HDOH has no PO43- water quality standard for
open coastal waters (HDOH 2014). It should also be noted that a ‘medium’ score in
nutrient concentrations exceeds HDOH standards for open coastal waters wet criteria.
Once each indicator was assigned a level (1-3) based on its measured value and
our scoring system (Table 4), its level was multiplied by a weight factor (1-3), with the
most reliable sewage indicators having the greatest weight. The greatest weight (weight =
17
3) was given to C. perfringens and δ 15N in macroalgal tissue because these indicators are
more specific to sewage pollution, more integrative measurements of environmental
conditions, and do not fluctuate as much as Enterococcus counts and nutrient
concentrations (Fung et al 2007; Dailer et al. 2010; Viau et al. 2011; Yoshioka et al.
2016). Enterococcus received a medium weight (weight = 2) as HDOH uses this FIB to
assess marine recreational water safety specifically for sewage pollution, but not the
highest weight because counts fluctuate over short time scales (min to h) and have other
sources, like soils, in tropical areas (Hardina & Fujioka 1991; Byappanahalli & Fujioka
1998; Byappanahalli & Fujioka 2004). Nutrient concentrations received the lowest
weight (weight = 1) since sewage pollution is known to increase nutrient concentrations,
but nutrients can also come from other sources within the watershed and concentrations
can vary over short time scales (Lapointe et al. 1990; David et al. 2013; Nelson et al.
2015). The equation for deriving the overall sewage pollution score for each station was:
(C. perfringens level x 3) + (δ15N macroalgae level x 3) + (Enterococcus level x 2) +
(NO3-+NO2
- level x 1) + (NH4+ level x 1) + (TDP level x 1). Sewage pollution score
categories were: ‘low’ = 11-15, ‘medium’ = 16-20, ‘high’ = 21-30.
The stations with highest pollution sewage scores were station 7 (score =30) and 4
(score =30) (Fig. 7). Note, that based on dye tracer data, these two stations are known
locations of OSDS leakage. Station 3 (score = 27), another location of known OSDS
leakage, had the third highest pollution score. These results provide confirmation of the
effectiveness of our sewage pollution score in identifying hotspots of sewage pollution.
Overall, 13 stations fell in the high category, two were medium, and one was low. This
integrated approach identified sewage hotspots along the Puakō coastline, and locations
where it is critical for homes to remove their cesspools and employ better sewage
treatment technology. This map also provides information to the community on areas
where community members may want to limit water exposure during recreational
activities until sewage treatment is improved.
Conclusion
Globally, coral reefs are declining from multiple stressors, with sewage pollution
being one of the most devastating (Wear & Vega Thurber 2015). FIB counts and nutrient
18
concentrations were high and variable along the Puakō shoreline, and δ 15N in macroalgal
tissue were within the determined range for sewage at Puakō. Dye tracer tests further
confirmed locations of sewage pollution and provided information on the time of travel
of the sewage from the homes to the shoreline. However, data from different sewage
indicators were not always in agreement with one another on the intensity and locations
of sewage pollution. Hence, a novel sewage pollution score was developed to use these
indicators to identify hotspots of sewage pollution. This approach allows a holistic
visualization of locations of sewage pollution. With sewage becoming a growing global
threat in nearshore waters, being able to effectively assess sewage pollution is crucial for
both human and marine ecosystem health. A multi-indicator approach for detecting
sewage pollution and this sewage pollution scoring system will allow other coastal
communities to assess their water quality and take appropriate management actions to
improve safety of recreational waters users and coastal ecosystem health.
19
Station Enterococcus
Clostridium
perfringens δ15N
1 18 ± 8b
[9-43] 2 ± 1 [0-4]
6.38 ± 0.15a-c
[6.03-6.65]
2 74 ± 25ab
[37-143] 2 ± 1 [0-4]
7.54 ± 0.18a-c
[7.04-7.90]
3 349 ± 162ab
[37-739] 5 ± 2 [1-10]
10.55 ± 0.15ab
[10.37-11.00]
4 237 ± 178ab
[47-770] 6 ± 2 [3-13]
11.88 ± 0.32a
[11.27-12.78]
5 1107 ± 861ab
[94-3674] 4 ± 1 [0-6]
7.21 ± 2.01a-c
[1.29-10.26]
6 1051 ± 570ab
[72-2546] 6 ± 2 [2-10]
7.77 ± 1.30a-c
[4.15-10.18]
7 170 ± 32ab
[104-257] 12 ± 5 [3-27]
8.35 ± 0.74a-c
[6.48-9.80]
8 738 ± 603ab
[62-2544] 7 ± 1 [3-10]
5.48 ± 0.37a-c
[5.06-6.58]
9 216 ± 120ab
[27-563] 3 ± 0 [2-3]
4.79 ± 0.53bc
[3.85-6.18]
10 122 ± 30ab
[66-202] 5 ± 1 [2-7]
4.54 ± 0.70c
[3.57-6.56]
11 315 ± 107ab
[15-495] 8 ± 3 [2-14]
6.02 ± 0.30a-c
[5.59-6.91]
12 676 ± 251ab
[121-1323] 5 ± 2 [2-8]
6.43 ± 0.65a-c
[4.90-8.04]
13 2777 ± 1806a
[17-7985] 2 ± 1 [0-3]
7.74 ± 1.92a-c
[4.80-13.12]
14 80 ± 36ab
[24-185] 10 ± 5 [1-20]
5.94 ± 0.47a-c
[5.08-7.28]
15 454 ± 132ab
[180-816] 9 ± 3 [2-13]
4.24 ± 0.51c
[3.62-5.77]
16 699 ± 554ab
[17-2338] 3 ± 1 [0-6]
4.23 ± 0.44c
[3.53-5.50]
Table 1. Average ± SE and [range] of Enterococcus (MPN/100 mL), C.
perfringens (CFU/100 mL), and δ15N in macroalgal tissue (‰) for shoreline stations at Puakō, Hawaiʻi. Superscript letters indicate significant groupings from One-way ANOVA and post-hoc Tukey’s tests. α = 0.05; n = 4.
20
N Source n δ15N in NO3- NO3
- + NO2- NH4
+ PO43-
Cesspools 3 10.45 ± 0.58 20.7 ± 10.5 6370 ± 806 378.6 ± 16.6 Soil 3 2.13 ± 2.37 6366.7 ± 3682.5 595 ± 93 193.6 ± 141.6
Ocean 2 3.02 ± 0.79 1.4 ± 0.1 3 ± 1 0.1 ± 0.1 High elevation
groundwater wells 3 4.76 ± 0.43 93.9 ± 4.4 5 ± 1 2.5 ± 0.2
Low elevation groundwater wells
7 7.03 ± 0.50 130.1 ± 6.7 5 ± 1 2.5 ± 0.5
Shoreline 3 11.95 ± 1.13 133.9 ± 64.7 n/a n/a
Table 2. Average ± SE of δ15N - NO3- (‰) and NO3
- + NO2-, PO4
3-, and NH4+ concentrations
(µmol/L) of N sources collected in the Puakō watershed. (n = sample size)
21
Station NO3- + NO2
- NH4+ TDN PO4
3- TDP H4SiO4 Salinity
1 27.87 ± 4.09b-e
[18.10-36.79] 20.83 ± 0.15 [0.78-1.23]
41 ± 7c-f
[25-58] 0.44 ± 0.04fg
[0.33-0.51] 0.70 ± 0.12fg
[0.51-1.04] 133 ± 23a-c
[87-195] 27.58 ± 1.44a-c
[23.63-30.37]
2 149.94 ± 12.79ab
[129.62-187.09] 0.49 ± 0.11 [0.18-0.72]
159 ± 13ab
[139-195] 2.24 ± 0.24a-d
[1.62-2.73] 2.86 ± 0.26a-e
[2.21-3.45] 581 ± 155ab
[187-876] 7.12 ± 0.61e [5.77-8.70]
3 137.12 ± 35.39a-c
[36.22-190.37] 1.95 ± 0.30 [1.04-2.29]
154 ± 39a-c
[41-217] 3.81 ± 0.92ab
[1.34-5.37] 4.28 ± 0.72ab
[2.42-5.09] 377 ± 124a-c
[112-646] 16.26 ± 3.96b-e [9.50-25.73]
4 196.05 ± 28.14a
[125.66-263.07] 1.3 ± 0.05 [1.24-1.47]
221 ± 26a
[153-267] 7.42 ± 1.11a
[4.12-9.0] 8.25 ± 1.36a
[4.45-10.84] 501 ± 113ab
[172-683] 15.25 ± 2.30c-e
[9.10-20.20]
5 46.92 ± 8.73a-e
[23.44-65.52] 1.32 ± 0.16 [0.86-1.57]
70 ± 12a-f
[42-87] 1.34 ± 0.17b-f
[0.90-1.71] 1.74 ± 0.28b-f
[0.90-2.13] 179 ± 41a-c
[85-278] 24.98 ± 2.35a-d
[19.70-31.07]
6 26.78 ± 11.48de
[2.50-54.16] 1.22 ± 0.10 [1.03-1.46]
44 ± 16d-f
[23-86] 0.66 ± 0.21e-g
[0.25-1.17] 0.85 ± 0.22fg
[0.25-1.26] 95 ± 43c
[22-219] 30.77 ± 2.31a [24.53-35.53]
7 134.56 ± 54.94a-d
[42.27-285.74] 1.69 ± 0.65 [0.46-2.90]
131 ± 43a-d
[53-241] 3.08 ± 0.44a-c
[2.12-3.83] 3.41 ± 0.50a-c
[2.19-4.51] 447 ± 132ab
[164-804] 21.98 ± 0.97a-d [19.87-24.03]
8 39.15 ± 14.53c-e
[0.99-67.10] 2.40 ± 0.97 [1.00-0.33]
59 ± 19b-f
[12-99] 0.70 ± 0.23e-g
[0.52-1.07] 1.01 ± 0.21e-g
[0.56-1.55] 253 ± 83a-c
[31-416] 20.60 ± 4.90a-d
[14.10-35.17]
9 69.74 ± 9.06a-e
[47.81-91.92] 1.00 ± 0.33 [0.89-1.77]
85 ± 7a-e
[74-105] 1.37 ± 0.13b-f [1.15-1.73]
1.80 ± 0.17b-f
[1.48-2.30] 342 ± 90a-c
[219-609] 15.28 ± 2.31cd [8.53-18.53]
10 56.72 ± 17.48a-e
[11.59-94.94] 0.95 ± 0.27 [0.47-1.51]
73 ± 19b-f
[20-106] 1.14 ± 0.31c-g
[0.34-1.84] 1.48 ± 0.16b-f
[1.18-1.84] 354 ± 76a-c
[129-445] 15.03 ± 3.60de [4.90-21.90]
11 16.52 ± 1.21de
[14.08-18.73] 0.96 ± 0.30 [0.18-1.45]
29 ± 4ef
[23-41] 0.49 ± 0.04e-g
[0.40-0.58] 0.76 ± 0.22fg
[0.25-1.33] 108 ± 27bc
[53-173] 28.30 ± 0.93ab
[26.07-30.60]
12 35.80 ± 4.37a-e
[25.62-46.59] 1.34 ± 025 [0.78-1.88]
46 ± 5b-f
[34.2-55.6] 0.99 ± 0.11c-g
[0.40-1.31] 1.26 ± 0.29c-g
[0.91-2.11] 260 ± 105a-c
[112-568] 24.50 ± 0.96a-d [22.57-27.13]
13 34.89 ± 4.73a-e [22.54-44.18]
1.21 ± 0.19 [0.73-1.56]
49 ± 7b-f
[35-67] 1.64 ± 0.28b-e
[0.91-2.29] 1.89 ± 0.17b-f
[1.66-2.38] 207 ± 23a-c
[167-267] 23.96 ± 2.00a-d [19.90-28.27]
14 89.08 ± 5.48a-d [75.93-101.22]
1.15 ± 0.29 [0.64-1.54]
101 ± 7a-d
[84-117] 2.61 ± 0.17a-c
[2.22-2.98] 2.91 ± 0.27a-d
[2.35-3.61] 652 ± 174a
[359-1018] 6.43 ± 0.63e [5.33-8.07]
15 13.37 ± 2.80e
[5.73-19.24] 1.07 ± 0.17 [0.75-1.44]
22 ± 3f
[15-27] 0.39 ± 0.09g
[0.16-0.55] 0.57 ± 0.21g
[0.25-1.12] 120 ± 24a-c
[52-158] 29.94 ± 0.70a [28.67-31.27]
16 38.5 ± 7.2a-e
[17.35-47.44] 0.63 ± 0.31 [0.18-1.51]
46 ± 4c-f
[34-52] 0.81 ± 0.13d-g
[0.45-1.09] 1.14 ± 0.30d-g
[0.60-1.99] 323 ± 86a-c
[142-552] 17.13 ± 3.44b-e [7.94-24.53]
Table 3. Average ± SE and [range] of NO3- + NO2
-, NH4+, TDN, PO4
3-, TDP, H4SiO4 concentrations (µmol/L), and salinity for shoreline stations at Puakō, Hawaiʻi. Superscript letters indicate significant groupings from One-way ANOVA and post-hoc Tukey’s tests. α = 0.05; n = 4.
22
Sewage Indicator Weight Factor
Low (1)
Medium (2)*
High (3) Reference
C. perfringens 3 0 – 10 11 - 100 101 – 505+ Fung et al. 2007 δ15N in macroalgae 3 +2 - +7 -5 - +1.9 +7 - +20 Wiegner et al. 2016
Enterococcus 2 0 - 35 36 - 104 105+ HDOH 2014 NO3
- + NO2- 1 0 – 0.4 0.5 – 1 1.1 – 1.8+ HDOH 2014
NH4+ 1 0 – 0.25 0.26 – 0.61 0.61 – 1.07+ HDOH 2014
TDP 1 0 – 0.7 0.8 – 1.3 1.4 – 1.9+ HDOH 2014
Table 4. Indicators (FIB = CFU/100 mL, δ15N = ‰, and nutrients = µmol/L) used to evaluate water quality health along the Puakō coastline. Sewage indicators were ranked (low = 1, medium = 2, high = 3), multiplied by a weight factor, and summed for a final sewage pollution score. * “Medium” nutrient concentration ranks exceed HDOH standards.
23
Figure 1. FIB, nutrients, δ15 N in macroalgae and water quality paraemeters were taken from 16 stations along the Puakō coastline, Hawaiʻi, USA (black circles). Dye tracer tests were conducted in proximity to the stations with squares. Nitrogen sources sampled included cesspools, high and low elevation groundwater wells, and soil.
Up-mountain
Cesspool
High elevation groundwater wells
Low elevation groundwater wells
Soil
24
0
2
4
6
8
10
12
14
δ15
N(‰
)
*
*Ap <0.0001
1
10
100
1000
10000
Ente
roco
ccus
(MP
N/1
00
mL)
*Bp =0.04
0
5
10
15
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16C. p
erfr
ing
ens
(CF
U/1
00
mL)
Cp =0.06
Figure 2. Average ± SE of sewage indicators: (A) δ15 N (‰) of residential macroalgae, (B) Enterococcus, and (C) C. perfringens at 16 shoreline stations at Puakō, Hawaiʻi. Single sample maximum for Enterococcus (104 CFU/100 mL) and a recommended standard for C. perfringens (5 CFU/100 mL) is indicated by black lines. Dashed black line respresents non-point sewage pollution (10 CFU/100 mL). Results from One-way ANOVA and Tukey’s tests are shown on figure, with * indicating significant differences (α = 0.05).
Station
25
-2
0
2
4
6
8
10
12
14
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
δ1
5N
(‰
)
Station
Soil
Ocean
High elevation GW
Low elevation GW
Sewage
Fertilizer
Figure 3. Average ± SE δ15 N (‰) of residental macroalgae found at 16 stations in Puakō, Hawaiʻi. Background areas represent (average ± SE) δ15 NO3
- of the N sources (fertilizer, soil, ocean, high elevation groundwater wells, low elevation groundwater wells, and sewage) measured as part of this study. Fertilizer values are from previous sewage study on Hawaiʻi Island (Wiegner et al. 2016).
26
Figure 4. Mixing plots of nutrient concentrations along the Puakō shoreline, Hawaiʻi: (A) NO3
- + NO2-, (B) NH4+, (C) TDN, (D) PO4
3-, (E) TDP, and (F) H4SiO4. Line represents theoretical mixing line, connecting freshest and saltiest shoreline samples. Groundwater samples were only analyzed for NO3
- + NO2-, NH4
+, and PO43- (orange triangles).
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0 10 20 30
TD
P (
µm
ol/L
)
E
0
50
100
150
200
250
300
0 10 20 30
TD
N (
µm
ol/L
)
C
0.0
2.0
4.0
6.0
8.0
10.0
0 10 20 30
PO
43-(µ
mol
/L)
D
0
200
400
600
800
1000
1200
0 10 20 30
H4S
iO4
(µm
ol/L
)
F
0.0
50.0
100.0
150.0
200.0
250.0
300.0
350.0
0 10 20 30
A
0
1
2
3
4
5
6
0 10 20 30
NH
4+ (u
mol
/L)
B
NO
3- + N
O2-
(µm
ol/L
)
Salinity (‰)
Salinity (‰)
Station 3
Station 4
Station 7
Other stations
GW Wells
27
Figure 5. Correlation between C. perfringens and NH4+along the
Puakō shoreline, Hawaiʻi.
0
1
2
3
4
5
6
0 5 10 15 20 25 30
NH
4+(µ
mol
/L)
C. perfringens (CFU/100 mL)
p =0.02r =0.28
28
0.0
50.0
100.0
150.0
200.0
250.0
300.0
350.0
NO
3-+
NO
2-(µ
mol
/L)
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
PO
43-(µ
mol
/L)
0
50
100
150
200
250
300
0 2 4 6 8 10 12 14
TD
N (
µm
ol/L
)
δ15 N (‰)
Figure 6. Correlations between δ15 N in macroalgal tissue and nutrient concentrations: (A) NO3
- + NO2-, (B) PO4
3-, and (C) TDN along the Puakō shoreline, Hawaiʻi.
A p <0.001 r =0.47
B p <0.001 r =0.61
C p <0.001 r =0.37
29
Figure 7. Sewage pollution scores for 16 stations along Puakō’s shoreline were created based on established and recommended water quality standards and literature values for sewage indicators (FIB, δ15 N in macroalgae, and nutrients). Sewage pollution score catergories are: Low = 11 - 15; Medium = 16 - 20; High = 21 - 30.
Medium
Low
High
30
Literature Cited
Abbott IA, Huisman JM (2004) Marine green and brown algae of the Hawaiian Islands. Honolulu, Hawaii: Bishop Museum Press, Honolulu, Hawaii
Abbott IA (1999) Marine red algae of the Hawaiian Islands. Bishop Museum Press, Honolulu, Hawaiʻi
Bisson JW, Cabelli VJ (1979) Membrane filter enumeration method for Clostridium perfringens. Appl Environ Microbiol 37:55-66
Byappanahalli M, Fujioka R (2004) Indigenous soil bacteria and low moisture may limit but allow faecal bacteria to multiply and become a minor population in tropical soils. Water Sci Technol 50:27-32
Byappanahalli M, Fujioka R (1998) Evidence that tropical soil environment can support the growth of Escherichia coli. Water Sci Technol 38:171-174
Converse RR, Griffith JF, Noble RT, Haugland RA, Schiff KC, Weisberg SB (2012) Correlation between quantitative PCR and culture- based methods for measuring Enterococcus spp. over various temporal scales at three California marine beaches. Appl Environ Microbiol 78:1237-1242
Costanzo SD, Udy J, Longstaff B, Jones A (2005) Using nitrogen stable isotope ratios (δ15N) of macroalgae to determine the effectiveness of sewage upgrades: Changes in the extent of sewage plumes over four years in Moreton Bay, Australia. Mar Pollut Bull 51:212-217
Couch C, Garriques J, Barnett C, Preskitt L, Cotton S, Giddens J, Walsh W (2014) Spatial and temporal patterns of coral health and disease along leeward Hawai’i Island. Coral Reefs 33:693-704
Dailer M, Smith J, Smith C (2012) Responses of bloom forming and non-bloom forming macroalgae to nutrient enrichment in Hawaiʻi, USA. Harmful Algae 17:111-125
Dailer ML, Ramey HL, Saephan S, Smith CM (2012) Algal δ15N values detect a wastewater effluent plume in nearshore and offshore surface waters and three-dimensionally model the plume across a coral reef on Maui, Hawaiʻi, USA. Mar Pollut Bull 64:207-213
Dailer ML, Knox RS, Smith JE, Napier M, Smith CM (2010) Using δ15N values in algal tissue to map locations and potential sources of anthropogenic nutrient inputs on the island of Maui, Hawai‘i, USA. Mar Pollut Bull 60:655-671
David A, Tournoud M, Perrin J, Rosain D, Rodier C, Salles C, Bancon-Montigny C, Picot B (2013) Spatial and temporal trends in water quality in a Mediterranean temporary river impacted by sewage effluents. Environ Monit Assess 185:2517-2534
31
Davies C, Long J, Donald M, Ashbolt N (1995) Survival of fecal microorganisms in marine and freshwater sediments. Appl Environ Microbiol 61:1888-1896
Derse E, Knee KL, Wankel SD, Kendall C, Berg CJ, Paytan A (2007) Identifying sources of nitrogen to Hanalei Bay, Kauai, utilizing the nitrogen isotope signature of macroalgae. Environ Sci Technol 41:5217-5223
Dollar S, Grigg R (2004) Anthropogenic and natural stresses on selected coral reefs in Hawaiʻi: A multidecade synthesis of impact and recovery. Pac Sci 58:281-304
Dudley B, Hughes F, Ostertag B (2014) Groundwater availablity mediates the ecosystem effects of an invasion of Prosopis pallida. Ecol Appl 24:1954-1971
Fong P, Donohoe R, Zedler J (1994) Nutrient concentration in tissue of the macroalga Enteromorpha as a function of nutrient history: An experimental evaluation using field microcosms. Mar Ecol Prog Ser 106:273-281
Fujioka R, Roll B, Byappanahalli N (1997) Appropriate recreational water quality standards for Hawaii and other tropical regions based on concentrations of Clostridium perfringens. Proceedings of the Water Environment Federation 4:405-411
Fujioka R, Hashimoto H, Siwak E, Young R (1981) Effect of sunlight on survival of indicator bacteria in seawater. Appl Environ Microbiol 41:690-696
Fujita RM (1985) The role of nitrogen status in regulating transient ammonium uptake and nitrogen storage by macroalgae. J Exp Mar Biol Ecol 92:283-301
Fung DYC, Fujioka R, Vijayavel K, Sato D, Bishop D (2007) Evaluation of Fung double tube test for Clostridium perfringens and easyphage test for f‐specific RNA coliphages as rapid screening tests for fecal contamination in recreational waters of Hawaii. Journal of Rapid Methods & Automation in Microbiology 15:217-229
Gallaher T, Merlin M (2010) Biology and impacts of Pacific Island invasive species. 6. Prosopis
pallida and Prosopis juliflora (algarroba, mesquite, kiawe) (Fabaceae). Pac Sci 64:489-526
Gaspar, E.1987. Modern trends in tracer hydrology, vols. I and II. CRC Press, Inc, Boca Raton, FL.
Hardina C, Fujioka R (1991) Soil: The environmental source of Escherichia
coli and Enterococci in Hawaii’s streams. Environ Toxicol Water Qual 6:185-195
Hayes T, Hourigan T, Jazwinski S, Johnson S, Parrish J, Walsh D (1982) The coastal resources, fisheries and fishery, ecology of Puako, west Hawaii. Hawaii Cooperative Fishery Research Unit 81-1:1-245
32
HDAR (2008) South Kohala reefs in dire straits. DAR. Handout.
HDOH (2014) State of Hawaii water quality monitoring and assessment report. HDOH CWA §305(b)
Hsing-Juh L, Wu C, Kao S, Kao W, Meng P (2007) Mapping anthropogenic nitrogen through point sources in coral reefs using δ15N in macroalgae. Mar Ecol Prog Ser 335:95-109
Hunt C (2006) Ground-water nutrient flux to coastal waters and numerical simulation of wastewater injection at Kihe, Maui, Hawaii. USGS; Scientific Investigations Report 5283:1-69
Hunter C, Evans C (1995) Coral reefs in Kaneohe Bay, Hawaii: Two centuries of western influence and two decades of data. Bull Mar Sci 57:501-515
Kendall C, McDonnell J (1998) Tracing nitrogen sources and cycling in catchments. In: Anonymous Tracing Nitrogen Sources and Cycling in Catchments. Elsevier Science B.V., Amsterdam, p 519-576
Knee K, Layton B, Street J (2008) Sources of nutrients and fecal indicator bacteria to nearshore water on the north shore of Kaua‘i (Hawai‘i USA). Estuar Coasts 31:607-622
Lapointe B, Barile P, Wynne M, Yentsch C (2005) Reciprocal Caulerpa invasion: Mediterranean native Caulerpa ollivieri in the Bahamas supported by human nitrogen enrichment. Aquat Invaders 16:1-5
Lapointe B, O'Connell J, Garrett G (1990) Nutrient couplings between on-site sewage disposal systems, groundwaters, and nearshore surface waters of the Florida Keys. Biogeochemistry 10:289-307
Littler MM, Littler DS (1980) The evolution of thallus form and survival strategies in benthic marine macroalgae: Field and laboratory tests of a functional form model. Am Nat 116:25-44
Maiga Y, Wethe J, Denyigba K, Ouattara AS (2009) The impact of pond depth and environmental conditions on sunlight inactivation of Escherichia coli and Enterococci in wastewater in a warm climate. Can J Microbiol 55:1364-1374
Minton D, Conklin E, Weiant P, Wiggins C (2012) 40 years of decline on Puakō’s coral reefs - A review of historical and current data (1970-2010) Honolulu, The Nature Conservancy, p 1-140
Nelson C, Donahue M, Dulaiova H, Goldberg S, La Valle F, Lubarsky K, Miyano J, Richardson C, Silbiger N, Thomas F (2015) Fluorescent dissolved organic matter as a multivariate
33
biogeochemical tracer of submarine groundwater discharge in coral reef ecosystems. Mar Chem 177:232-243
Nnane DE, Ebdon JE, Taylor HD (2011) Integrated analysis of water quality parameters for cost- effective faecal pollution management in river catchments. Water Res 45:2235-2246
Ochoa‐Izaguirre MJ and Soto‐Jiménez MF (2015) Variability in nitrogen stable isotope ratios of macroalgae: Consequences for the identification of nitrogen sources. J Phycol 51:46-65
Parsons ML, Walsh WJ, Settlemier CJ, White DJ, Ballauer JM, Ayotte PM, Osada KM, Carman B (2008) A multivariate assessment of the coral ecosystem health of two embayments on the lee of the island of Hawai‘i. Mar Pollut Bull 56:1138-1149
Pastorok R, Bilyard G (1985) Effects of sewage pollution on coral-reef communities. Mar Ecol Prog Ser 21:175-189
Paytan A, Shellenbarger GG, Street JH, Gonneea ME, Davis K, Young MB, Moore WS (2006) Submarine groundwater discharge: An important source of new inorganic nitrogen to coral reef ecosystems. Limnol Oceanogr 51:343-348
Peterson F, Oberdorfer J (1985) Uses and abuses of wastewater injection wells in Hawaii. Pac Sci 39 2:230-240
Pinto B (1999) Characterization of faecal streptococci as indicators of faecal pollution and distribution in the environment. Lett Appl Microbiol 29:258-263
Reich, C., E.A. Shinn, C. Hickey and A.B. Tihansky. 2001. Tidal and meteorological influences on shallow marine groundwater flow in the upper Florida Keys. In: J. Porter and K.C Porter (Editors) The Everglades, Florida Bay, and coral reefs of the Florida Keys, An Ecosystem Handbook. CRC Press.
Savage C (2005) Tracing the influence of sewage nitrogen in a coastal ecosystem using stable nitrogen isotopes. Ambio 34:145-150
Schott E (2010) Mauna Lani, DW ʻĀina Leʻa and Puakō/Wailea Bay wastewater assessment. Professional Report to the Puakō Community Association Phase One. p 1-47
Shibata T, Solo-Gabriele H, Fleming LE, Elmir S (2004) Monitoring marine recreational water quality using multiple microbial indicators in an urban tropical environment. Water Res 38:3119-3131
Shibata T, Solo-Gabriele H, Sinigalliano CD, Gidley ML, Plano LRW, Fleisher JM, Wang JD, Elmir SM, He G, Wright ME, Abdelzacher AM, Ortega C, Wanless D, Garza AC, Kish J, Scott T, Hollenback J, Backer LC, Fleming LE (2010) Evaluation of conventional and
34
alternative monitoring methods for a recreational marine beach with nonpoint source of fecal contamination. Environ Sci Technol 44:8175-8181
Shuval H (2003) Estimating the global burden of thalassogenic diseases: Human infectious diseases caused by wastewater pollution of the marine environment J Water Health 1:53-64
Sutherland K, Ritchie K (2004) White pox disease of the Caribbean elkhorn coral Acropora
palamata. Coral Health and Disease 3:289-300
Swart PK, Evans S, Capo T, Altabet MA (2014) The fractionation of nitrogen and oxygen isotopes in macroalgae during the assimilation of nitrate. Biogeosciences 11:6147–6157
Umezawa Y, Miyajima T, Yamamuro M, Kayanne H, Koike I (2002) Fine- scale mapping of land- derived nitrogen in coral reefs by δ15N in macroalgae. Limnol Oceanogr 47:1405-1416
Unesco (1981). Background papers and supporting data on the practical salinity scale 1978. Unesco technical papers in marine science 37
US Census Bureau (2000). Profile of General Demographic Characteristics: 2000
USEPA (2003) Bacterial water quality standards for recreational waters (freshwater and marine waters) status report. Epa-823-r-03-008 :1-32
USEPA (2013) Cesspools in Hawaiʻi. http://www.epa.gov/region9/water/groundwater/uic-hicesspools.html (accessed 2 Aug 2014)
Vega Thurber RL, Burkepile DE, Fuchs C, Shantz AA, Mcminds R, Zaneveld JR (2014) Chronic nutrient enrichment increases prevalence and severity of coral disease and bleaching. Global Change Biol 20:544-554
Viau E, Goodwin K, Yamahara K, Layton B, Sassoubre L, Burns S, Tong H, Wong S, Lu Y, Boehm A (2011) Bacterial pathogens in Hawaiian coastal streams associations with fecal indicators, land cover, and water quality. Water Res 45:3279-3290
Wang Y, Wilson JM, Vanbriesen JM (2015) The effect of sampling strategies on assessment of water quality criteria attainment. J Environ Manage 154:33-39
Wear S, Vega Thurber R (2015) Sewage pollution: Mitigation is key for coral reef stewardship. Ann N Y Acad Sci :1-16
Whittier R, El-Kadi A (2014) Human health and environmental risk ranking of on-site sewage disposal systems for the Hawaiian islands of Kauai, Molokai, Maui, and Hawaii. http://health.hawaii.gov/wastewater/files/2015/09/OSDS_NI.pdf
35
Wiegner TN, Mokiao-Lee A, Johnson EE (2016) Identifying nitrogen sources to thermal tide pools in Kapoho, Hawai'i, U.S.A, using a multi- stable isotope approach. Mar Pollut Bull 103:63-71
Xue D, Botte J, De Baets B, Accoe F, Nestler A, Taylor P, Van Cleemput O, Berglund M, Boeckx P (2009) Present limitations and future prospects of stable isotope methods for nitrate source identification in surface- and groundwater. Water Res 43:1159-1170
Yoshioka RM, Kim CJS, Tracy AM, Most R, Harvell CD (2016) Linking sewage pollution and water quality to spatial patterns of Porites lobata growth anomalies in Puako, Hawaii. Mar Pollut Bull 104:313-321
Zambrano L, Contreras V, Mazari-Hiriart M, Zarco-Arista A (2009) Spatial heterogeneity of water quality in a highly degraded tropical freshwater ecosystem. Environ Manage 43:249-263
Zimmerman M (December 22, 2010) Sewage spill forced into Kailua Bay where president Obama and family are staying. http://www.hawaiireporter.com/sewage-spill-forced-into-kailua-bay-where-president-obama-and-family-are-staying/123 (accessed 7 Jun 2013).
36
Spatial distribution of sewage pollution on a Hawaiian coral reef
Chapter 2
37
Abstract
Sewage pollution has been shown to affect both human health and nearshore ecosystems,
and is especially a concern in tropical regions with coral reefs. Puakō, located on Hawaiʻi Island,
is one area of concern because sewage pollution has been detected along on its shoreline;
however, the spatial distribution of sewage pollution offshore in surface and benthic waters is
unknown. This study examined the spatial extent of sewage pollution using algal bioassays and a
combination of sewage indicators (fecal indicator bacteria (FIB), stable nitrogen isotopes (δ15N)
in macroalgal tissue, nutrients). FIB counts and nutrients were spatially variable in both surface
and benthic waters, with shoreline concentrations exceeding water quality standards. δ15N in
macroalgal tissue, along the shoreline were the most enriched and within the range for sewage.
However, sewage indicators were not always in agreement on the location and intensity of
pollution. To assess water quality with regards to these indicators, a sewage pollution score was
created, allowing for locations of sewage pollution hotspots to be identified. This approach for
identifying sewage pollution hotspots is valuable for other coastal communities with documented
sewage pollution so that appropriate management actions can be taken to improve water quality,
and reduce human and ecosystem health hazards.
38
Introduction
Worldwide, sewage pollution impacts coastal waters. Sewage can enter coastal waters
from accidental spills from sewage treatment plants, injection wells, and leaks from cesspools
and septic tanks (Hunt 2006, Knee et al. 2008, Zimmerman 2010). Sewage pollution introduces
pathogens (bacteria and viruses), nutrients, hydrocarbons, toxins, and endocrine disruptors to
nearshore waters, threatening human health and coastal ecosystems (USEPA 2012 1990, Wear &
Thurber 2015). Over 120 million gastroenteric illnesses have been reported globally within a
year in relation to sewage-contaminated coastal waters (Shuval 2003). Exposure to sewage,
either from ingestion, inhalation, and direct contact, may result in hepatitis and skin and urinary
tract infections (Pinto et al. 1999, USEPA 2011). The United States Environmental Protection
Agency (USEPA) estimates that nationwide 61% of small communities (<10,000 people) use
cesspools or septic tanks for sewage disposal (USEPA 2012), which can contaminate
groundwater and nearshore waters with untreated sewage (Bruno et al. 2003; Friedlander et al.
2005; Paytan et al. 2006; Fung et al. 2007; Boehm et al. 2010).
Coral reefs are one coastal ecosystem particularly sensitive to sewage pollution. Sewage
affects the prevalence and severity of coral disease (Sutherland et al. 2010; Vega et al. 2014). In
addition, high levels of sewage-derived nutrients can alter coral calcification rates, reshape
microbial communities, and affect coral symbionts (Wear & Thurber 2015). Nutrient enrichment
has also been shown to trigger elevated levels of algal growth and phytoplankton blooms,
resulting in eutrophication (Hunter & Evans 1995; Lapointe et al. 2005; Paytan et al. 2006;
Rabalais et al. 2009). This increase in algal growth can cause detrimental ecological shifts on
coral reefs, such that macroalgal density increases and fish abundance and coral cover decreases
(Reopanichkul et al. 2009; Wear & Thurber 2015).
Measuring levels of fecal indicator bacteria (FIB) is a widely used method to detect
sewage pollution in coastal waters. FIB are used to assess the potential risk to human health in
recreational waters. Enterococcus is the most commonly used FIB in marine waters, but it has
been found to naturally thrive in tropical soils due to ideal temperatures and rich nutrient
substrates, and can multiply in seawater (Hardina & Fujioka 1991; Byappanahalli & Fujioka
1998; Byappanahalli & Fujioka 2004). Thus, in tropical states like Hawaiʻi, Clostridium
perfringens has been adopted as an additional FIB for sewage pollution because it does not
39
multiply in aerobic seawater (Davies et al. 1995; Fung et al. 2007). In addition, Enterococcus has
been shown to be an indicator of recent sewage contamination, whereas C. perfringens reflects
continuous inputs (Curiel-Ayala et al. 2012), and thus, both FIB should be used concurrently to
assess sewage pollution.
In addition to FIB, some studies measure stable nitrogen (N) isotopes (δ15N) in
macroalgal tissue to detect sewage in coastal environments (Umezawa et al. 2002; Savage 2005;
Hsing-Jun Lin et al. 2007;Wiegner et al. 2016). Stable N isotopes are used to determine N
sources to coastal waters as N sources can have distinct isotopic signatures, and macroalgal
tissues reflect the N they uptake due to their minimal discrimination between 14N and 15N during
uptake (Savage 2005). For example, treated sewage gets highly enriched due to denitrification
with δ15N values for NO3- ranging from +7‰ to +20‰, differing from fertilizers (-5‰ to +5‰),
soil NO3- (-10‰ to +15‰), seawater NO3
- (+2‰ to +7‰), and atmospheric N2 (0‰) (reviewed
in Wiegner et al. 2016). Often, the macroalgal species chosen in sewage-related studies as
bioindicators are opportunistic ones whose growth is stimulated by increases in N (Littler &
Littler 1980; Dailer 2010; Dailer 2012). Ulva fasciata is a common species used because it is
commonly found along the shoreline and has rapid nutrient uptake rates (Abbott & Huisman
2004). More recently, the use of plant tissues to evaluate sewage pollution (macroalgal
bioassays) was established to examine sewage within water bodies, and to encompass offshore
transport patterns (Costanzo et al. 2001; Costanzo et al. 2005.) This technique has been used to
map improvements in water quality following sewage treatment plant upgrades (Costanzo et al.
2005), nearshore sewage plumes (Dailer et al. 2010; Dailer et al. 2012), nutrient inputs from
local fish farms (Garcia-Sanz et al. 2010; Garcia-Sanz et al. 2011), and to detect cruise ship
wastewater inputs (Kaldy et al. 2011).
Usually, studies focus on a single indicator to assess sewage pollution in coastal waters,
with only a few studies using two or more (Shibata et al. 2004; Knee et al. 2008; Dailer et al.
2010; Lenart-Boron et al. 2015; Yoshioka et al. 2016). Using a single indicator to assess sewage
pollution may be misleading as different indicators can vary greatly over time and space, and
different factors affect different indicators differently. For example, studies that have evaluated
both FIB Enterococcus and C. perfringens to monitor sewage pollution found that Enterococcus
was more likely to exceed recreational water quality standards in more locations compared to C.
40
perfringens (Griffin et al. 1999; Shibata et al. 2014; Weisz 2014). Thus, beach closures and
advisories would be more likely if Enterococcus levels were used instead of C. perfringens. In
addition, monitoring δ15N in macroalgal tissue has been shown to be highly variable depending
on contributing N sources in the area (Ochua-Izaquirre & Soto-Jimenez 2015). Therefore,
measuring more than a single sewage indicator is vital for accurately assessing sewage pollution.
Furthermore, most studies that have sampled FIB and δ15N in macroalgae have been restricted to
the shoreline (Bonkosky et al. 2009; Dailer et al. 2010; Shibata et al. 2010; Yoshioka et al.
2016). However, sewage can be transported offshore or enter coastal waters through benthic
seeps. For example, elevated Enterococcus levels that are indicative of sewage have been
detected nearly a mile offshore (Weisz 2014). Sewage has also been detected in both offshore
and benthic waters using δ15N in macroalgal tissues (Dailer et al. 2012; Dailer et al. 2010). These
studies highlight that determining the spatial distribution of sewage offshore and in the benthos is
critical to improving water quality for humans and ecosystems.
Hawaiʻi is one state dealing with human and coral reef health threats from sewage
pollution. Hawaiʻi uses cesspools more widely than any other state in the United States (USEPA
2013), and a ban on new installments were only recently signed into state legislation (2015). The
greatest number of cesspools in Hawaiʻi State are located on Hawaiʻi Island, totaling ~49,000
(Whittier & El-Kadi 2014). Puakō, a coastal community located in the South Kohala district on
Hawaiʻi Island, is a known “area of concern” in the state. It is considered a high-risk area, where
the presence of onsite sewage disposal systems (OSDS) can affect nearshore environments
(Whittier & El Kadi 2014), and it has also been designated by Hawaiʻi State as a priority site for
site-based action due to its rich diversity of corals (Hayes et al. 1982; NOAA 2010). However,
over the last 40 years, decreases in coral coverage and fish abundances have been documented
(Minton et al. 2012). Increases in coral disease prevalence and severity have also been
documented (Couch et al. 2014; Yoshioka et al. 2016). These ecosystem changes are thought in
part to be due to sewage pollution.
To understand the spatial distribution of sewage pollution offshore of the Puakō
coastline, a combination of techniques were used. To assess the offshore spatial distribution of
sewage pollution in surface and benthic waters, FIB and nutrients were measured, and bioassays
quantifying δ15N in macroalgal tissues were deployed. Only a handful of macroalgal bioassay
41
studies have examined sewage pollution offshore in surface and benthic waters, and even fewer
that used both FIB in combination with macroalgal bioassays (Dailer et al. 2010; Dailer et al.
2012; Yoshioka et al. 2016). The presence of sewage along the coast of Puakō has been
documented (Yoshioka et al. 2016; Chapter 1 thesis); however, the spatial extent of sewage
offshore in surface and in the benthos was not determined. Sewage pollution is a global concern
and assessing the presence of sewage in coastal waters is vital to human and ecosystem health
(Wear & Thurber 2015). This study assessed whether sewage was present offshore in surface and
benthic waters. Other coastal communities worldwide will be able to use this approach to
determine the spatial distribution of sewage offshore, allowing them to accurately monitor water
quality and develop solutions and management tools to improve environmental conidtions for
both human and coastal ecosystems health.
Methods
Site Description
The Puakō community has more than 200 homes along the coast (<0.001 km) and relies
heavily on cesspools (~47) and septic tanks (~139) for wastewater disposal (Schott 2010). The
Puakō area is mostly comprised of basalt from the Mauna Loa Volcano flow in 1859, and thus,
infiltration is high at the ground’s surface due to the porous nature of the substratum. Puakō
receives 250 – 750 mm of rainfall annually and groundwater discharge ranges from 2083 to 2730
L m-1 h-1 (Paytan et al. 2006). The hydrological connectivity between cesspools and the
nearshore environment in Puakō ranges from 9 h to 3 d (Chapter 1 thesis). The shoreline
substratum consists mostly of turf algae, basalt with little limestone, and coral coverage that
increases with increasing distance from the shoreline; however, turf algae is the most dominant
benthic species in both shallow and deeper benthic regions.
Study Design
To determine the spatial extent of sewage pollution offshore, as well as possible inputs
from benthic seeps that could directly impact the coral reefs, water was sampled for FIB and
nutrients. Additionally, the green macroalga, Ulva fasciata, was deployed during bioassays for
δ15N analysis at five stations (Fig. 1). These stations encompassed three zones (shoreline, bench,
and slope) and two depths (surface and benthic) (Fig. 1). Benthic zones were chosen based on
physiography features. The bench zone was ~7 m deep, and ~196 m from the shoreline. The
42
slope zone was ~15 m in depth, and ~267 m from the shoreline. The bench and slope zones were
~65 m apart. Collection of water samples and algal cage deployments were conducted in June
and July 2015. There was one sample collection and cage deployment per month.
FIB & Nutrient Analyses
Water samples were collected once during each deployment at all zones and depths, and
analyzed for FIB, nutrients, and salinity. FIB Enterococcus was analyzed using the Enterolert
MPN method (IDEXX Laboratories Inc) and C. perfringens was quantified using a membrane
filtration technique (Bisson & Cabelli 1979). Water samples for nutrient analysis were filtered
through pre-combusted (500°C for 6 hours), 0.7-µm pore size GF/F filters (WhatmanTM) and
stored frozen until analysis at the University of Hawaiʻi at Hilo (UHH) Analytical Laboratory.
Nutrients in water samples were analyzed on a Pulse TechniconTM II autoanalyzer using standard
methods and reference materials (NIST; HACH 307-49, 153-49, 14242-32, 194-49). These
samples were analyzed for NO3- + NO2
- [Detection Limit (DL) 0.07 µmol L-1, USEPA 353.4],
NH4+ [DL 0.36 μmol L-1, USGS I-2525], PO4
3- [DL 0.03 μmol L-1, Technicon Industrial Method
155-71 W], total dissolved phosphourous (TDP) [DL 0.5 μmol L-1, USGS I-4650-03], and
H4SiO4 [DL 1 μmol L-1, USEPA 366]. Total dissolved nitrogen (TDN) was analyzed by high-
temperature combustion, followed by chemiluminescent detection of nitric oxide (DL 5 μmol L-
1, Shimadzu TOC-V, TNM-1). Salinity was assessed using an YSI Pro 2030.
Ulva fasciata Deployments & Analysis
Ulva fasciata (Chlorophyta) was collected locally and acclimated to low nutrient water to
ensure N stores within the thalli were depleted prior to deployments. A sample of the U. fasciata
was collected and preserved as a voucher for identification. Preliminary studies determined that
Instant-Ocean (33‰-35‰) was the most suitable low-nutrient medium, with an acclimation
period of 3 d for U. fasciata (Fig. 2), during which water was changed every 24 h. Thalli of U.
fasciata were deployed in netted plastic cages (mesh size ~ 5 mm x 5 mm) creating a protective
barrier from herbivores, while allowing sunlight and water movement within the cages. These
cages were incubated in a grid-like pattern at the three zones for seven days offshore of the
Puakō coastline (Fig. 1). Within each zone, excluding the shoreline, three cages containing U.
fasciata were placed at two depths: ~1 m near the surface with subsurface buoys (surface) and
three above the reefs (~1 m) secured by weights (benthic) (Fig. 1). Approximately 4 g of U.
43
fasciata were rinsed with reagent-grade water, spun-dried, and weighed before being placed in
cages. In addition, to determine if there was a difference between deployed U. fasciata and
nearby in situ macroalgae, wild algae were also collected at all benthic zones during the June
deployment.
Following deployments, U. fasciata and adjacent wild algae were retrieved, rinsed with
reagent-grade water, dried to 60° C until a constant weight was achieved, ground, and
homogenized using a Wig-L-Bug grinding mill. For stable isotope analysis, 2 mg of the
macroalgal tissues were folded in 4x6 mm tin capsules. Macroalgal tissues were analyzed for
δ15N using a Thermo-FinniganTM Delta V Advantage isotope ratio mass spectrometer (IRMS)
with a Conflo III interface and a CostechTM ECS 4010 Elemental Analyzer located at the UHH
Analytical Laboratory. Data were normalized to USGS standard NIST 1547. Isotopic signatures
were expressed as standard (δ) values, in units of parts per mil (‰), and calculated as: [(Rsample –
Rstandard) / Rstandard] x 1000, where R = 15N/14N.
Statistical Analyses
To examine differences among sewage indicators among zones, general linear models
(GLM) were used. One model examined surface water patterns extending from the shoreline to
offshore. The second model examined benthic patterns from the shoreline to offshore. To
determine differences between depths, a nested GLM was used, where depth was nested within
zones to account for variability at each station. In addition, to determine differences between
initial and final δ15N in U. fasciata, a one-way analysis of variance (ANOVA) was used with
initial values compared with shoreline, as well as surface and benthic offshore zones. δ15N in U.
fasciata was also compared to N sources that were taken in a companion study (Chapter 1
thesis). To determine if differences in δ15N existed between adjacent wild macroalgae and
deployed U. fasciata, a two-sample t-test was used. Data were tested for normality and equal
variances; if assumptions for parametric analyses were not met, log, log + 1, and rank
transformations were used (Potvin & Roff 1993). Statistical analyses were conducted using
Minitab 16TM (α=0.05).
44
Results
Surface and Benthic Waters Spatial Patterns
Enterococcus counts were similar among zones in surface waters (p =0.29; Fig. 3);
however, counts did differ significantly among benthic zones (p =0.04; Fig. 3). The greatest
differences in the benthos were detected between shoreline (302 MPN/100 mL ± 306) and slope
(35 MPN/100 mL ± 22) zones, which were almost an order of magnitude different. In contrast,
C. perfringens differed significantly among zones in both surface (p =0.01) and benthic waters (p
<0.01). In surface waters, the largest differences were detected between shoreline (8 CFU/100
mL ± 4) and slope (2 CFU/100 mL ± 1) zones (Fig. 3B). Shoreline C. perfringens counts were
also significantly higher (8 CFU/100 mL ± 4) compared to benthic bench (1 CFU/100 mL ± 1)
and benthic slope (1 CFU/100 mL ± 0.22) waters (Fig. 3E). Nutrient concentrations (NO3- +
NO2-, NH4
+, TDN, PO43-, TDP, and H4SiO4) were highest on the shoreline in both surface (p
<0.02) and benthic (p <0.01) waters (Table 1). Nutrient concentrations among zones in surface
and benthic waters were similar between bench and slope zones. Salinity concentrations also
varied among zones in both surface (p<0.01) and benthic waters (p<0.01), with the shoreline
having the freshest (lowest) value (18.52 ± 3.08). δ15N in U. fasciata varied significantly by zone
in both surface (p =0.01) and benthic waters (p<0.01) (Fig. 3). Shoreline values were the highest
(5.61‰ ± 0.77), followed by slope (surface = 4.32‰ ± 0.34, benthic = 4.18‰ ± 0.34), and bench
(surface = 3.92‰ ± 0.46, benthic = 3.71‰ ± 0.29). Both δ15N for surface and benthic U. fasciata
samples fell within the δ15N - NO3- range for soil, seawater, and low elevation groundwater at all
zones determined in a companion study (Chapter 1 thesis) (Fig. 4).
Surface vs. Benthic Waters
No differences in depth were detected among sewage indicators: Enterococcus (p >0.43),
C. perfringens (p >0.10), nutrient concentrations [NO3- + NO2
- (p >0.71), NH4+ (p >0.17), TDN
(p >0.29), PO43- (p >0.66), and TDP (p >0.32)], and δ15N in U. fasciata (p >0.08). H4SiO4
concentrations did vary, with the greatest differences detected between surface waters at the
bench and benthic waters at the slope (p <0.01). Salinity levels between surface and benthic
waters were also similar (p >0.90).
45
Initial vs. Post Deployment δ15N values
Initial (3.49‰ ± 0.44) and post-deployment δ15N U. fasciata values differed (p <0.01),
with the greatest differences occurring at the shoreline (5.69‰ ± 0.23)(Fig. 5). The slope zone in
surface waters (4.32‰ ± 0.17) and benthic waters (4.19‰ ± 0.21) showed smaller differences in
initial and post deployment δ15N, followed by the bench zone in surface waters (4.08‰ ± 0.45)
and benthic waters (3.83‰ ± 0.22).
Wild Macroalgae vs. Deployed U. fasciata
Adjacent wild collected macroalgae had δ15N values similar to those deployed in the
benthic zones; however, differences were detected at the shoreline. δ15N in wild algae in the
bench zone ranged from -0.57‰ to +4.02‰, with an average of +2.90‰ ± 1.96. Deployed U.
fasciata in the bench zone ranged from +3.23‰ to +4.27‰, with an average of +3.83‰ ± 0.49.
In the slope zone, δ15N in wild algae ranged from +3.48‰ to +8.92‰ (+6.09‰ ± 2.31) and
deployed U. fasciata ranged from +3.50‰ to +4.78‰ (+4.19‰ ± 0.48). However, no significant
difference in δ15N values between wild algae and deployed U. fasciata were observed (p =0.58).
At the shoreline, wild algae ranged from +5.07‰ to +10.18‰, with an average of +7.75‰ ±
1.25. Deployed shoreline U. fasciata ranged from +3.37‰ to +7.27‰, with an average of
+5.61‰ ± 1.08. Significant differences between wild macroalgae and deployed U. fasciata were
also detected (p =0.01). In both cases at the shoreline, the highest δ15N values were observed at
the same station (station 2).
Discussion
FIB counts are often spatially variable within a water body, with highest levels generally
observed closer to the shoreline (Paul et al. 1995; Griffen et al. 1999; Shibata et al. 2004;
Bonkosky et al. 2009; Lisle et al. 2014). Similarly, our FIB counts were spatially variable in both
surface and benthic waters. Enterococcus concentrations in the surface waters were similar
across zones. However, in the benthic waters, Enterococcus concentrations were significantly
higher along the shoreline compared to bench and slope zones. Clostridium perfringens in both
surface and benthic waters were greatest at the shoreline, with the greatest differences being
between shoreline and slope zones. The spatial variability observed at Puakō is similar to FIB
spatial differences reported for other water bodies (Shibata et al. 2004). For example, in the
Florida Keys, Enterococcus and C. perfringens levels were highly variable according to distance
46
from the shoreline, with the highest counts being reported nearshore (Paul et al. 1995). Similar
patterns in Hawai’i have also been observed, where high counts of Enterococcus were
consistently seen over a 1.6 km from shore, and C. perfringens counts were highest nearshore
(Weisz 2014).
FIB concentrations at the shoreline exceeded established and recommended recreational
water quality standards in Hawaiʻi. Enterococcus counts in surface waters at all three zones
(inclusive of SE) exceeded the Hawaiʻi Department of Health (HDOH) single sample maximum
(104 CFU/100mL). At the shoreline, counts were three times greater than HDOH’s single sample
maximum. The benthic waters at the shoreline and bench zones also exceeded the HDOH
maximum, suggesting these waters may be impaired. Additionally, in surface and benthic waters,
C. perfringens counts at the shoreline exceeded recommendations of 5 CFU/100mL for marine
recreational waters, and fell above the Fung/Fujioka scale for non-point source sewage
contamination levels (> 10 CFU/100mL) (Fujioka et al. 1997; Fung et al. 2007). This suggests
that the shoreline may be polluted by sewage.
Similar to FIB, nutrient concentrations were observed to be spatially variable among
zones and between depths. Nutrient concentrations were highest at the shoreline and decreased
with increasing distance from shore. In the Guangzhou Sea, China, a similar inshore to offshore
nutrient gradient was seen in an area with significant sewage inputs (Peng & Liangmin 2010).
Similar spatial nutrient patterns have been observed in Hilo Bay, Hawaiʻi, a known impaired
water body (Weisz 2014). Furthermore, in a similar sewage study on Hawaiʻi Island, in which
nutrient concentrations were assessed in tide pools adjacent to homes with cesspools, a
comparable decreased level onshore to offshore was observed (Wiegner et al. 2016). Compared
to other studies conducted in Hawaiʻi, NO3- + NO2
- values were 10x greater than those values
observed on Kauaʻi in Hanalei Bay, an area known for sewage contaminated groundwater (Knee
et al. 2008). In addition, our nutrient data were consistently higher than Hilo Bay and Hanalei
Bay, with some nutrient concentrations nearly five times greater at the shoreline. In addition,
Puakō nutrient concentrations (NO3- + NO2
-, NH4+, PO4
3+) were the highest that have been
reported, and in some cases, 60x greater than those previously reported for Puakō (Knee et al.
2010). Salinity levels indicated the freshest seeps occurred near the shoreline where nutrients are
highest, suggesting that freshwater may be a source of nutrients.
47
Elevated shoreline δ15N macroalgal values in comparison to offshore sites have been
documented in many locations impacted by sewage (Yoshioka et al. 2016; Dailer et al. 2010;
Chapter 1 thesis). Lower values offshore probably reflect smaller direct inputs of sewage near
benthic seeps or dilution of nearshore inputs. In this study, the shoreline had the highest δ15N
macroalgal values, followed by slope and bench zones in both surface and benthic waters. This is
similar to a study conducted in Maui, which found δ15N in macroalgal tissue to be most enriched
near a wastewater reclamation facility (WWRF), decreasing with increasing distance from the
shoreline and the WWRF (Dailer et al. 2010). However, our study indicated a slight gradient, in
which shoreline was significantly more enriched compared to bench and slope zones. This
similar gradient trend was seen in Hanalei Bay, Kauaʻi, highlighting the shoreline as a potential
hotspot compared to offshore sites (Derse et al. 2007). In contrast, a study detected fish farm
waste up to 2450 m offshore in Murcia, Spain, using macroalgal bioassays (Garcia-Sanz et al.
2011), suggesting fish farm waste was transported offshore. However, Puakō results indicate that
sewage transport offshore may not be occurring and is limited.
δ15N macroalgal values were found to be fall in the range for soil, seawater, and low
elevation groundwater. In addition, sewage transport offshore seems to be limited. The shoreline
δ15N values in U. fasciata may be underestimated by up to ~6‰ due to the high nutrient
concentration environment (NO3- concentrations >10 μmol/L). Under these conditions,
discrimination between 15N and 14N during nutrient uptake is enhanced (Swart et al. 2014). If this
occurred in our study, then all of our shoreline δ15N macroalgal values fall within local sewage
range. At several locations along the Puakō shoreline, the presence of sewage was confirmed
through dye tracer studies in which sewage hotspots were identified (Chapter 1 thesis). These are
thought to occur due to local geological features such as fractures and lava tubes. Input of
sewage at the shoreline could be a reason why nutrient concentrations are highly elevated
compared to offshore zones and other coastal locations within the state of Hawaiʻi.
Similarities between depths among sewage indicators were found in this study. A
previous study found sewage to rise from the benthos to surface waters, reflecting a stronger
δ15N signal in surface waters, and to follow a general direction away from the sewage source
(Dailer et al. 2012). However, during large wave mixing events, sewage was detected in the
benthos more so than the surface waters (Dailer et al. 2010). In contrast, no differences among
48
sewage indicators between depths were detected in our study. δ15N values in U. fasciata in
surface waters fell within the range of low elevation groundwater, which may be contaminated
with sewage given the numerous OSDS above the wells (Chapter 1 thesis). In contrast, the
benthic δ15N values in U. fasciata were within the range of a soil source. These differences could
be due in part to the macroalgae being exposed to different light levels and nutrient
concentrations and forms (NH4+) between two depths, as these parameters have been shown to
influence the growth rates in macroalgae, and therefore their capability to uptake N (Xu et al.
2014). The amount of available light at our deep depths could have also influenced the ability of
the macroalgae to uptake N. During deployments macroalgae were suspended near the benthos at
15 m in the slope zones. The ability of light to reach these deeper areas may not have been as
significant compared to those on the surface and may have influenced our δ15N values. However,
past studies have found Ulva spp. capable of taking up N as they were exposed to different N
forms and concentrations (Gartner et al. 2002; Dailer et al. 2010), which may indicate why δ15N
in U. fasciata values significantly increased from initial δ15N values compared to post-
deployment values.
Similar to chapter 1, our results found that sewage indicators provided conflicting
information on the intensity and location of sewage pollution among zones and depths. To date,
various sewage-related studies have examined a single sewage indicator to assess sewage
pollution such as FIB (Fujioka & Shizumura 1985; Fung et al. 2007), macroalgal bioassays using
δ15N (Kaldy 2011; Dailer et al. 2010; Dailer et al. 2012; Garcia-Sanz et al. 2010), and/or
combination of macroalgal bioassays using δ15N and FIB (Moynihan et al. 2012; Yoshioka et al.
2016), without assessing nutrient data. This study demonstrated that elevated levels of FIB,
nutrients, and δ15N in U. fasciata along the shoreline were indicative of sewage. However,
patterns among sewage indicators were not consistent. To assess potential sewage hotspots, a
sewage pollution score developed in a companion study was employed (Chapter 1 thesis). This
scoring system used the following sewage indicators according to water quality standard and/or
literature values for sewage contamination: Enterococcus (104 CFU/100 mL), C. perfringens (10
CFU/100 mL), δ15N in macroalgae tissue (> +7‰), and nutrients (NO3- + NO2
-, NH4+, TDP).
These indicators were then weighted based on their reliability as sewage indicators and used in
the following equation to determine a sewage pollution score for each station. Sewage pollution
49
score = (C. perfringens level x 3) + (δ15N macroalgae level x 3) + (Enterococcus level x 2) +
(NO3-+NO2
- level x 1) + (NH4+ level x 1) + (TDP level x 1) [Eq. 1]. Sewage pollution score
categories were: ‘low’ = 11-15, ‘medium’ = 16-20, ‘high’ = 21-30.
The shoreline zone fell within the low and medium sewage pollution scores, with stations
1 and 3 being potential hotspots (Fig. 6). This is similar to the results from Chapter 1 of this
study. However, in the first part of the study, majority of the shoreline stations fell within a high
sewage pollution score. Offshore, the majority of the stations fell within a low sewage pollution
score category with only stations 1 and 4 possibly having a sewage pollution contamination (Fig.
6). This holistic approach identified potential shoreline sewage pollution hotspots, and provides
additional information to the community about areas where onshore and offshore sewage
pollution may exist. In addition, our data in combination with coral health data can assist in
examining relationships between sewage and disease prevalence and severity.
Conclusion
An approach using multiple indicators is needed to detect sewage pollution in coastal
waters to accurately assess water quality conditions for both human and coral reef health. This is
the first study that employed the use of FIB, δ15N in macroalgae, and nutrients to assess the
spatial distribution of sewage offshore and in the benthos of a coral reef. However, because
sewage indicators didn’t agree on the locations and intensity of sewage pollution, a sewage
pollution score was utilized. From this, we were able to identify areas of concern in Puakō, due
to possible sewage pollution. By combining our sewage pollution score with coral disease and
prevalence we may be able to establish a link between them. Using multiple techniques to assess
sewage pollution will provide other coastal communities around the world an application to
holistically determine spatial patterns of sewage pollution and to identify areas in need of
management action to improve water quality for human and ecosystems health.
50
Figure 1. Location of water sample collection (for FIB and nutrients) and algal cage deployments (for δ15N in U. fasciata) in Puakō, Hawaiʻi. Water and macroalgal samples were taken at three zones (shoreline, bench, deep) to determine the spatial extent of sewage pollution in surface and benthic waters offshore, Diagram of algal cage deployment design is shown in lower right corner of figure.
Benthic Weight
Buoy
Surface 1 m
1 m
51
0
1
2
3
4
5
6
7
1 2 3 7 14 21 28
Oce
an
Wate
r δ
15
N (
‰)
Time (days)
B - Ocean Water
0
1
2
3
4
5
6
7
1 2 3 7 14 21 28
Inst
an
t O
cean
δ1
5N
(‰
)
Time (days)
A - Instant Ocean
Figure 2. Changes in δ15N of U. fasciata during purging experiments with Instant Ocean (A) and ocean water (B).
52
Ben
thic
Wat
ers
0
1
2
3
4
5
6
7
Shoreline Bench Slope
δ15N
(‰
)
C p =0.01 a
b b
1
10
100
1000
Shoreline Bench Slope
En
tero
cocc
us*
(MP
N/1
00 m
L)
Ap =0.29
0
2
4
6
8
10
12
14
Shoreline Bench Slope
C. P
erfr
ing
ens
(CF
U/1
00 m
L)
B p =0.01 a
ab b
Sur
face
Wat
ers
0
1
2
3
4
5
6
7
Shoreline Bench Slope
δ15N
(‰
)
Fp <0.01 a
b b
1
10
100
1000
Shoreline Bench Slope
Ente
roco
ccus*
(MP
N/1
00 m
L)
Dp =0.04 a
ab
b
0
2
4
6
8
10
12
14
Shoreline Bench Slope
C. p
erfr
ing
ens
(CF
U/1
00 m
L)
Ea
b b
p <0.01
Figure 3. Average ± SE of sewage parameters (A, D) Enterococcus (*logged scale), (B, E) C. perfringens, and (C, F) δ15N in U. fasciata
collected within three zones (shoreline, bench, slope) in both surface and benthic waters in Puakō. Black lines represent the HDOH single sample maximum for Enterococcus (104 CFU/100 mL) and Fujioka’s recommendation (1997) for C. perfringens in marine recreational waters (5 CFU/100mL). Dashed lines represent non-point source sewage contamination level of 10 CFU/100 mL for C. perfringens (Fung et al. 2007) Results from GLM and Tukey’s test are shown, with different letters indicating significant differences (α =0.05). FIB n =10. Sample size varied for δ15N in U. fasciata in both surface waters (shoreline, n =9; bench, n =6; slope, n =10) and benthic waters (shoreline, n =9; bench, n =8; slope, n =10).
53
-2
0
2
4
6
8
10
12
δ1
5N
(‰
)
Benthic Surface
Soil
Ocean
Low Elevation GW
High Elevation GW
Sewage
Fertilizer
Shoreline Bench Slope
Surface Benthic
Figure 4. Average ± SE δ15 N (‰) of U. fasciata deployed within three benthic zones (shoreline, bench, slope) in Puakō, Hawaiʻi. Background areas represent average ± SE of δ15 N – NO3
- of the N sources taken in a companion study at Puakō (Chapter 1 thesis) and fertilizer from another study on Hawaiʻi Island (Wiegner et al. 2016). Surface samples are represented by grey triangles and benthic samples by black circles.
54
Figure 5. Average ± SE δ15 N (‰) of U. fasciata pre-(initial) and post-deployments within three benthic zones (shoreline, bench, slope) and two depths (surface and benthic) in Puakō, Hawaiʻi. GLM was used and shared lettering indicates no significant differences in Tukey’s post hoc test. Sample size varied (initial, n =11; shoreline, n =5; surface bench, n =4; surface slope, n =5; benthic bench, n =5; benthic slope, n =5). α =0.05.
0
1
2
3
4
5
6
7
Initial Shoreline Bench Slope Bench Slope
δ1
5N
(‰
)
p <0.01
b
a
ab abb
b
Surface Benthic
55
Figure 6. Map of sampling stations at Puakō with their sewage pollution scores. Larger circles represent benthic stations and smaller circles represent surface stations. Scores were created based on established and recommended water quality standards and literature values for sewage indicators. Sewage pollution scores represent the following categories: Low = 11 – 15; Medium = 16 – 20; High = 21 – 30.
Low
Medium
High
56
Zone NO3- + NO2
- NH4+ TDN PO4
3- TDP H4SiO4 Salinity Shoreline 66.87 ± 11.47a
[11.59 – 139.72] 1.52 ± 0.16a [0.18 – 3.05]
73 ± 11a [21 – 121]
1.67 ± 0.22a [0.47 – 2.56]
1.98 ± 0.22a [0.70 – 3.25]
439 ± 74a [154 – 617]
18.52 ± 3.08a
[3.78 – 29.63] Surface
Bench 1.43 ± 0.26b [0.83 – 1.84]
0.57 ± 0.14b [0.18 – 1.56]
10 ± 1b [8 – 12]
0.14 ± 0.03b [0.02 – 0.27]
0.64 ± 0.13b [0.25 – 1.23]
7 ± 3b
[1 – 21] 33.26 ± 1.11b
[29.95 – 34.47] Slope 1.23 ± 0.18b
[0.40 – 2.14] 0.38 ± 0.11b [0.18 – 1.06]
9 ± 1b [7 – 13]
0.12 ± 0.02b [0.02 – 0.24]
0.59 ± 0.11b [0.25 – 0.96]
5 ± 1b
[1 – 11] 34.24 ± 0.41b
[33.75 – 34.62] Benthic
Bench 1.10 ± 0.13b
[0.53 – 2.06] 0.50 ± 0.12b [0.18 – 1.23]
10 ± 1b [7 – 13]
0.18 ± 0.05b [0.02 – 0.49]
0.58 ± 0.11b [0.25 – 0.94]
2 ± 1b
[1 – 5] 33.55 ± 0.95b
[31.03 – 35.0] Slope 1.57 ± 0.51b
[1.01 – 6.09] 1.10 ± 0.53ab [0.18 – 5.58]
9 ± 1b [7 – 13]
0.24 ± 0.11b [0.02 – 1.13]
0.94 ± 0.29b [0.25 – 3.25]
1 ± 0b
[1 – 1] 34.46 ± 0.30b
[34.22 – 34. 85]
Table 1. Average ± SE and [range] of nutrient concentrations (μmol/L) and salinity for surface and benthic water samples among zones (shoreline, bench, slope) in Puakō, Hawaiʻi. A GLM was used to determine differences among zones and between depths, and superscript letters indicate grouping from post hoc Tukey’s test. α = 0.05; n = 10.
57
Literature Cited
Abbott IA, Huisman JM (2004) Marine green and brown algae of the Hawaiian Islands. Honolulu, Hawaii: Bishop Museum Press, Honolulu, Hawaii
Bisson JW, Cabelli VJ (1979) Membrane filter enumeration method for Clostridium membrane filter enumeration method for Clostridium perfringens. Appl Environ Microbiol 37:55–66
Boehm A, Yamahara K, Walters S, Layton B, Keymer D, Thompson S, Knee K, Rosner M (2011) Dissolved inorganic nitrogen, soluble reactive phosphorous, and microbial pollutant loading from tropical rural watersheds in Hawai‘i to the coastal ocean during non-storm conditions. Estuar Coasts 34:925–936
Bonkosky M, Hernandez-Delgado EA, Sandoz B, Robledo IE, Norat-Ramrez J, Mattei H (2009) Detection of spatial fluctuations of non-point source fecal pollution in coral reef surrounding waters in southwestern Puerto Rico using PCR-based assays. Mar Pollut Bull 58:45–54
Bruno JF, Petes LE, Harvell CD, Hettinger A (2003) Nutrient enrichment can increase the severity of coral diseases. Ecol Lett 6:1056–1061
Byappanahalli M, Fujioka R (1998) Evidence that tropical soil environment can support the growth of Escherichia coli. Water Sci Technol 38:171–174
Byappanahalli M, Fujioka R (2004) Indigenous soil bacteria and low moisture may limit but allow faecal bacteria to multiply and become a minor population in tropical soils. Water Sci Technol 50:27–32
Costanzo SD, O’Donohue MJ, Dennison WC, Loneragan NR, Thomas M (2001) A new approach for detecting and mapping sewage effects. Mar Pollut Bull 42:149–156
Costanzo SD, Udy J, Longstaff B, Jones A (2005) Using nitrogen stable isotope ratios (δ15N) of macroalgae to determine the effectiveness of sewage upgrades: Changes in the extent of sewage plumes over four years in Moreton Bay, Australia. Mar Pollut Bull 51:212–217
Couch CS, Garriques JD, Barnett C, Preskitt L, Cotton S, Giddens J, Walsh W (2014) Spatial and temporal patterns of coral health and disease along leeward Hawaiʻi Island. Coral Reefs 33:693–704
Curiel-Ayala F, Quinones-Ramirez EI, Pless RC, Gonzalez-Jasso E (2012) Comparative studies on Enterococcus, Clostridium perfringens and Staphylococcus aureus as quality indicators in tropical seawater at a Pacific Mexican beach resort. Mar Pollut Bull 64:2193–2198
58
Dailer ML, Knox RS, Smith JE, Napier M, Smith CM (2010) Using δ15N values in algal tissue to map locations and potential sources of anthropogenic nutrient inputs on the island of Maui, Hawai’i, USA. Mar Pollut Bull 60:655–671
Dailer ML, Ramey HL, Saephan S, Smith CM (2012) Algal δ15N values detect a wastewater effluent plume in nearshore and offshore surface waters and three-dimensionally model the plume across a coral reef on Maui, Hawai’i, USA. Mar Pollut Bull 64:207–213
Dailer ML, Smith JE, Smith CM (2012) Responses of bloom forming and non-bloom forming macroalgae to nutrient enrichment in Hawai’i, USA. Harmful Algae 17:111–125
Davies CM, Long JAH, Donald M, Ashbolt NJ (1995) Survival of fecal microorganisms in marine and freshwater sediment. Appl Environ Microbiol 61:1888–1896
Derse E, Knee K, Wankel S, Kendall C, Paytan A (2010) Identifying sources of nitrogen to Hanalei Bay, Kauai, utilizing the nitrogen isotope signature of macroalgae. Environ Sci Technol 41:5217-5223
Friedlander A, Aeby G, Brown E, Clark A, Coles S, Dollar S, Hunter C, Jokiel P (2005) Main Hawaiian Islands. The state of coral reef ecosystems of the Main Hawaiian Island. p 222-269
Fujioka RS, Shizumura LK (2014) Clostridium perfringens, a reliable indicator of stream water quality. Water Pollut Control Fed 57:986–992
Fujioka R, Roll B, Byappanahalli M (1997) Appropriate recreational water quality standards for Hawaiʻi and other tropical regions based on concentrations of Clostridium perfringens. Proceedings of the Water Environment Federation 4:405-411
Fung DYC, Fujioka R, Vijayavel K, Sato D, Bishop D (2007) Evaluation of Fung double tube test for Clostridium perfringens and easyphage test for f‐specific RNA coliphages as rapid screening tests for fecal contamination in recreational waters of Hawaii. Journal of Rapid Methods & Automation in Microbiology 15:217-229
Garcia-Sanz T, Ruiz J, Perez M, Ruiz M (2011) Assessment of dissolved nutrients dispersal derived from offshore fish-farm using nitrogen stable isotope ratios (δ15N) in macroalgal bioassays. Estuar Coast Shelf Sci 91:361–370
Garcia-Sanz T, Ruiz-Fernández JM, Ruiz M, García R, González MN, Pérez M (2010) An evaluation of a macroalgal bioassay tool for assessing the spatial extent of nutrient release from offshore fish farms. Mar Environ Res 70:189–200
Gartner A, Lavery P, Smit AJ (2002) Use of δ15N signatures of different functional forms of macroalgae and filter-feeders to reveal temporal and spatial patterns in sewage dispersal. Mar Ecol Prog Ser 235:63–73
59
Griffin DW, Gibson CJ III, Lipp EK, Riley K, Paul JH III, Rose JB (1999) Detection of viral pathogens by reverse transcriptase PCR and of microbial indicators by standard methods in the canals of the Florida Keys. Appl Environ Microbiol 65:4118-4125
Harinda C, Fujioka R (1991) Soil: The environmental source of Escherichia coli and Enterococci in Hawaii’s streams. Environ Toxicol Water Qual An Int J 6:185–195
Hayes T, Hourigan T, Jazwinski S, Johnson S, Parrish J, Walsh D (1982) The coastal resources, fisheries and fishery, ecology of Puako, west Hawaii. Hawaii Cooperative Fishery Research Unit 81-1:1-245
HDAR (2008) South Kohala reefs in dire straits. DAR. Handout.
HDOH (2015) Cesspools in Hawaiʻi. http://health.hawaii.gov/wastewater (accessed 1 Jan 2016).
Hunt C (2006) Ground-water nutrient flux to coastal waters and numerical simulation of wastewater injection at Kihe, Maui, Hawaii. USGS; Scientific Investigations Report 5283:1-69
Hunter CL, Evans CW (1995) Coral reefs in Kaneohe Bay, Hawaii: Two centuries of western influence and two decades of data. Bull Mar Sci 57:501–515
Kaldy J (2011) Using a macroalgal δ15N bioassay to detect cruise ship waste water effluent inputs. Mar Pollut Bull 62:1762–1771
Knee KL, Layton BA, Street JH, Boehm AB, Paytan A (2008) Sources of nutrients and fecal indicator bacteria to nearshore waters on the north shore of Kauaʻi (Hawaiʻi, USA). Estuaries and Coasts 31:607–622
Knee K, Street JH, Grossman EG, Paytan A (2010) Nutrient inputs to the coastal ocean from submarine groundwater discharge in a groundwater-dominated system: Relation to land use (Kona coast, Hawaii, U.S.A.). Limnol Oceanogr 55:1105–1122
Lapointe BE, Barile PJ, Wynne MJ, Yentsch CS (2005) Reciprocal Caulerpa Invasion: Mediterranean native Caulerpa ollivieri in the Bahamas supported by human nitrogen enrichment. Coral Reefs 16:2–5
Lenart-Boroń A, Wolanin A, Jelonkiewicz Ł, Chmielewska-Błotnicka D, Żelazny M (2016) Spatiotemporal variability in microbiological water quality of the Białka River and its relation to the selected physicochemical parameters of water. Water Air Soil Pollut 227:1-12
Lin HJ, Wu CY, Kao SJ, Kao WY, Meng PJ (2007) Mapping anthropogenic nitrogen through point sources in coral reefs using δ15N in macroalgae. Mar Ecol Prog Ser 335:95–109
60
Lisle JT, Smith JJ, Edwards DD, McFeters GA (2004) Occurrence of microbial indicators and Clostridium perfringens in wastewater, water column samples, sediments, drinking water, and Weddell seal feces collected at McMurdo Station, Antarctica. Appl Environ Microbiol 70:7269–7276
Littler MM, Littler DS (1980) The evolution of thallus form and survival strategies in benthic marine macroalgae: field and laboratory tests of a functional form model. Am Nat 116:25–44
Looney EE, Sutherland KP, Lipp EK (2010) Effects of temperature, nutrients, organic matter and coral mucus on the survival of the coral pathogen, Serratia marcescens PDL100. Environ Microbiol 12:2479–2485
Minton D, Conklin E, Weiant P, Wiggins C (2012) 40 years of decline on Puakō’s coral reefs - A review of historical and current data (1970-2010) Honolulu, The Nature Conservancy, p 1-140
Moynihan MA, Baker DM, Mmochi AJ (2012) Isotopic and microbial indicators of sewage pollution from Stone Town, Zanzibar, Tanzania. Mar Pollut Bull 64:1348–1355
NOAA (2010) Hawaiian Archipelago’s: Coral reef management priorities. Hawaii’s coral reef management priorities. p 1-34
Ochoa-Izaguirre MJ, Soto-Jiménez M (2015) Variability in nitrogen stable isotope ratios of macroalgae: Consequences for the identification of nitrogen sources. J Appl Phycol 51:46–65
Paul JH, Rose JB, Jiang S, Kellogg C, Shinn EA (1995) Occurrence of fecal indicator bacteria in surface waters and the subsurface aquifer in Key Largo, Florida. Appl Environ Microbiol 61:2235-2241
Paytan A, Shellenbarger GG, Street JH, Gonneea ME, Davis K, Moore WS, Young MB (2006) Submarine groundwater discharge: An important source of new inorganic nitrogen to coral reef ecosystems. Limology Oceanogr 51:343–348
Peterson F and Oberdorfer J (1985) Uses and abuses of wastewater injection wells in Hawaii. Pac Sci 39 2:230-240
Pinto B, Pierotti R, Canale G, Reali D (1999) Characterization of “faecal streptococci” as indicators of faecal pollution and distribution in the environment. Lett Appl Microbiol 29:258–263
Potvin C, Roff D (1993) Distribution-free and robust statistical methods : Viable alternatives to parametric statistics. Ecology 74:1617–1628
61
Rabalais NN, Turner RE, Diaz RJ, Justic D (2009) Global change and eutrophication of coastal waters. ICES J Mar Sci 66:1528–1537
Reopanichkul P, Schlacher TA, Carter RW, Worachananant S (2009) Sewage impacts coral reefs at multiple levels of ecological organization. Mar Pollut Bull 58:1356–1362
Savage C, Source CS, Environment H, By P, Academy RS, Doi S (2005) Tracing the influence of sewage nitrogen in a coastal ecosystem using stable nitrogen isotopes. Ambio 34:145–50
Shibata T, Solo-Gabriele HM, Fleming LE, Elmir S (2004) Monitoring marine recreational water quality using multiple microbial indicators in an urban tropical environment. Water Res 38:3119–3131
Shuval H (2003) Estimating the global burden of thalassogenic diseases: Human infectious diseases caused by wastewater pollution of the marine environment. J Water Heal 1:53–64
Swart PK, Evans S, Capo T, Altabet MA (2014) The fractionation of nitrogen and oxygen isotopes in macroalgae during the assimilation of nitrate. Biogeosciences 11:6147–6157
Umezawa Y, Miyajima T, Yamamuro M, Kayanne H, Koike I (2002) Fine-scale mapping of land-derived nitrogen in coral reefs by δ15N in macroalgae. Limnol Oceanogr 47:1405–1416
USEPA (2011) Keeping raw sewage & contaminated stormwater out of the public’s water. https://www3.epa.gov/region02/water/sewer-report-3-2011.pdf
USEPA 2012 (1990) Water: U.S. census data on small community housing and wastewater disposal and plumbing practices. http://water.epa.gov/type/oceb/fact3.cfm (accessed 03 Jun 2015)
Vega Thurber RL, Burkepile DE, Fuchs C, Shantz AA, Mcminds R, Zaneveld JR (2014) Chronic nutrient enrichment increases prevalence and severity of coral disease and bleaching. Glob Chang Biol 20:544–554
Wear SL, Thurber RV (2015) Sewage pollution: Mitigation is key for coral reef stewardship. Ann N Y Acad Sci 1355:15–30
Wei P, Huang L (2010) Water quality and eutrophication in the Guangzhou Sea Zone of the Pearl River estuary. Chinese J Oceanol Limnol 28:113–121
Weisz C (2014) Spatial and temporal microbial pollution patterns in a tropical Hawaiian estuary between high and low river flow conditions. MS dissertation, University of Hawaiʻi at Hilo, HI
62
Whittier R and El-Kadi A (2014) Human health and environmental risk ranking of on-site sewage disposal systems for the Hawaiian islands of Kauai, Molokai, Maui, and Hawaii. http://health.hawaii.gov/wastewater/files/2015/09/OSDS_NI.pdf
Wiegner TN, Mokiao-Lee AU, Johnson EE (2016) Identifying nitrogen sources to thermal tide pools in Kapoho, Hawai’i, U.S.A, using a multi-stable isotope approach. Mar Pollut Bull 103:63–71
Xue D, Botte J, Baets B De, Accoe F, Nestler A, Taylor P, Cleemput O, Berglund M, Boeckx P (2009) Present limitations and future prospects of stable isotope methods for nitrate source identification in surface- and groundwater. Water Res 43:1159–1170
Yoshioka R, Kim CJS, Tracey A, Most R, Harvell CD (2016) Linking sewage pollution and water quality to spatial patterns of Porites lobata growth anomalies in Puako, Hawaii. Mar Pollut Bull 104:313-321
Zimmerman M (December 22, 2010) Sewage spill forced into Kailua Bay where president Obama and family are staying. http://www.hawaiireporter.com/sewage-spill-forced-into-kailua-bay-where-president-obama-and-family-are-staying/123 (accessed 7 Jun 2013).