Usage of this form is not approved for other program offices.
OMB Approval No. 0648-0384
Expiration Date: 3/31/2013
Format for Project Final Report
I. Report Title: Restoration within He‘eia Ahupua‘a: Effects on species diversity and water
quality
A.
Authors: Florence Thomas, Judy Lemus, Kimberly Peyton
B.
Organization: University of Hawaii at Manoa, Hawaii Institute of Marine Biology,
C.
Grant Number: NA09NOS4260242
D.
Date: 1 September 2010 to 31 August 2011.
II.
Executive Summary
Pollutants are a primary concern affecting the health of nearshore habitats in Hawaii. Invasive species are
a form of biological pollution which act as agents of change in coastal systems. In particular, the shifts in
species composition associated with invasions can influence water quality in wetlands, waterways and the
coastal ocean habitats. As a biologically rich tropical estuary, Kāne‘ohe Bay is an exceptional ecosystem
with extensive coral reefs, inland fresh- and brackish water ponds and streams fringed by expansive
wetlands. Regrettably both wetlands and coral reefs have been severely degraded by invasive species.
KĀKO‘O ‘ŌIWI has spearheaded a grassroots effort to restore the He‘eia ahupua‘a by shifting vegetation and
fish assemblages away from invasive dominated systems. We supported the efforts of KĀKO‘O ‘ŌIWI by
investigating how the functions of the wetlands and associated ponds changed as the ecosystem
transitions from an invasive species dominated to a restored habitat. Our work examined how restoration
efforts affected three primary functions of coastal wetlands: the transport of nutrients, sediments, and water
through the system. Additionally, we documented changes in species diversity and wildlife habitat, two
other primary wetland functions.
III.
Purpose
A.
Overarching goal(s) of the project:
Collect baseline data to evaluate whether degraded habitats characterized by invasive species or restored
native and traditional agricultural habitats are better at reducing terrestrial runoff into downstream coastal
habitats.
B.
Management problem addressed:
Understanding how restoration of freshwater coastal systems upstream can affect downstream processes
on coral reefs. Changes upstream can affect the reefs and this study provided valuable baseline data on
how upstream restoration will translate to coral reef health.
C.
Objectives of the project.
Objective #1: Collect baseline data by documenting the species occurring the He‘eia inland freshand brackish water ponds and wetlands.
Objective #2: Trace biological responses before and during the shifts from an invasive species
dominated system to one under restoration.
Objective #3: Record chemical and physical responses before and during the shifts from an
invasive species dominated system to one under restoration.
Objective #4: Train interns and community volunteers to monitor and maintain their restoration
efforts.
IV.
Approach
A.
Detailed description of the work that was performed.
Objective #1: Collect baseline data by documenting the species occurring the He‘eia inland freshand brackish water ponds and wetlands.
Species lists of vascular plants, fish and macroalgae were developed from surveys conducted from
October 2010 to June 2011. Each of the four extant open water ponds was surveyed twice to
account for any seasonal differences. Additionally, kalo loi were surveyed as both the areas that
had been restored to taro within the past two years and new loi that were opened up during the
coarse of this project. All observed wetland vascular plants were identified to species using the
Hawai`i Wetland Field Guide: An Ecological and Identification Guide to Wetlands and Wetland
Plants of the Hawaiian Islands (Erickson and Puttock (2006)). Fish were sampled using a variety of
techniques including traps, hook and line, case net and siene net. Fish were determined to the
lowest taxonomic unit possible (because of hybridization of introduced species) using multiple
sources.
Objective #2: Trace biological responses before and during the shifts from an invasive species
dominated system to one under restoration.
Maps of vegetative canopy types along open water margins were created using a GPS set on
tracking mode. The GPS recorded a position every minute and waypoints were recorded to mark
changes in dominant vegetation. ArcGIS software was used to generate maps that projected open
water area traces along with dominant vegetation on to base imaginary (aerial) from USGS. This
approach was to compare open water area during this funding with historic photographs and to
compare vegetation changes before and immediately following a storm event in June 2011.
Additional data collected included:
 water depth
 habit of dominant species: fully submergent, partially submergent or entirely emergent and
rooted, floating mat or free floating
 dominant species canopy height: above water and below water
 other species present in the dominant canopy
Objective #3: Record chemical and physical responses before and during the shifts from an
invasive species dominated system to one under restoration.
Coastal ponds
To compare water quality associated with dominant vegetation types two YSI 600XLM sondes
were used. One sonde was dedicated as the control and positioned at fixed site located in the
center of a pond with no emergent or submergent vegetation nearby. A second sonde was
deployed for four days each in four different vegetation canopies: Job’s tears grass (Coix
lachryma), Neke fern (Cyclosorus interruptus), California grass (Urochloa mutica) and Aka akia
sedge (Schoenoplectus tabernaemontani). Sondes recorded temperature, salinity, water level,
oxygen and pH every ten minutes. Three replicate time periods were used. Additionally, water
samples were collected at four discrete depths (10, 25, 50 75 cm) within each canopy and the
control site. A set of water samples was sent to University of Washington Oceanography Technical
Services for nutrient sample analyses using a Technicon Auto Analyzer II (AAII) for PO4, Si(OH)4,
NO3, NO2 and NH4. A second set of water samples was filtered through preweighed and precombusted 47 mm Glass microfiber filters (Whatman GF/F Catalog Number: 1825-047) to measure
the amount of particles suspended in the water column. Water volume was adjusted per sample so
that it yielded between 2.5 to 200 mg of dried sample. Filters were dried at 60ºC to constant weight
and weighed to obtain total weight. Next filters were combusted at 550ºC for 4 hours and weighted
to obtain weight of inorganic matter. By subtracting weight of inorganic matter from total weight the
weight of organic matter was determined.
Taro loi.
To estimate how restoration efforts can affect the transport of nutrients during storm events through
the taro loi water samples were taken at four locations: stream enter, loi enter, loi exit and stream
exit. Varying amounts of water from a stream (at discretion of the farm manager) are diverted into a
taro loi that was recently restored (covered in California grass 2 years ago). The sampling strategy
followed water movement patterns currently being practiced by the farm manager as follows: water
is feed (stream enter) into the loi through pipes (loi enter) and eventually exits the loi via earthen
ditches (loi exit) back into the stream (stream exit). A 25-year storm event (flash flood) occurred 4-5
June 2011. Water samples were taken on 3 June (pre-storm), 5 June (storm), 7-12 June and 30
June (post-storm) and sent to University of Washington Oceanography Technical Services for
nutrient sample analyses using a Technicon Auto Analyzer II (AAII) for PO4, Si(OH)4, NO3, NO2
and NH4. Sondes were also deployed in taro loi.
Objective #4: Train interns and community volunteers to monitor and maintain their restoration
efforts.
Dr. Lemus met regularly with Kako‘o ‘Oiwi outreach staff to discuss educational
applications of the monitoring techniques that were being developed by Drs. Peyton and
Thomas. Interns from The Nature Conservancy, as well as a few Kako‘o ‘Oiwi
volunteers, also received training in several monitoring
techniques, including GPS waypointing, vegetation
identification and mapping, vegetation coverage transects, and
salinity, temperature, dissolved oxygen, and pH
measurements. Over the course of the monitoring sessions,
modifications of the procedures were tested to determine their
effectiveness with volunteer groups. Developing protocols
that resulted in useful and reliable data, while allowing
relatively untrained laypersons to participate in the research,
were the primary considerations. Four parameters became
the basis for the community-based monitoring program:
Dissolved oxygen; Total suspended solids; Vegetation
mapping; and Plankton surveys.
Fig. 1 Kyrie Yonehiro learning to take water samples for
total suspended solids and nutrient analysis.
B.
Project management: List individuals and/or organizations actually performing
the work and how it was done.
Summary of Management: Drs. Thomas and Lemus oversaw the completion of the work while Dr.
Peyton oversaw and supervised field research. Dr. Thomas’ lab at HIMB lent labor and equipment
to the project and provided data direction for data analysis.
Objective #1: Collect baseline data by documenting the species occurring the He‘eia inland freshand brackish water ponds and wetlands.
Field surveys were preformed by Kimberly Peyton, Lauren Valentino, Kaleo Hurley (HIMB), Kyrie
Puaoi, Colton Jones (Kāko‘o ‘Ōiwi), and Brad Wong, Nahaku Kalei (The Nature Conservancy).
Data input was done by Lauren Valentino, Kaleo Hurley (HIMB). Identification of specimens was
supervised by Kimberly Peyton (HIMB).
Objective #2: Trace biological responses before and during the shifts from an invasive species
dominated system to one under restoration.
Field surveys were preformed by Kimberly Peyton, Lauren Valentino, Kaleo Hurley (HIMB), Kyrie
Puaoi, Colton Jones (Kāko‘o ‘Ōiwi), and Brad Wong, Nahaku Kalei (The Nature Conservancy).
Data input was done by Lauren Valentino (HIMB). Maps using ArcGIS software were produced
by Nahaku Kalei (The Nature Conservancy).
Objective #3: Record chemical and physical responses before and during the shifts from an
invasive species dominated system to one under restoration.
Samples were collected and processed by Kimberly Peyton, (HIMB), Kyrie Puaoi (Kāko‘o ‘Ōiwi).
Sondes were deployed and maintained by Kimberly Peyton (HIMB). Data were managed by
Kimberly Peyton (HIMB). Thomas and Peyton conducted analysis and interpretation of data.
Objective #4: Train interns and community volunteers to monitor and maintain their restoration
efforts.
Dr. Lemus over saw educational and outreach program and worked with the assistance of
Daveyanne "Bootsie" Howard, and Jeffrey Liko Kaluhiwa.
V.
Findings
A.
Actual accomplishments and findings.
Objective #1: Collect baseline data by documenting the species occurring the He‘eia inland freshand wetlands.
Emergent vascular plants –
Four species of emergent plants formed extensive canopies along the margins of 4
extant open water areas surveyed:
Job’s tears grass (Coix lachryma)
Neke fern (Cyclosorus interruptus)
California grass (Urochloa mutica)
Aka akia sedge (Schoenoplectus tabernaemontani)
Additionally, nine species of emergent plants occurred within the extensive canopies and two
sedge species were associated with the restored kalo loi.
Submergent vascular plants –
No submergent vascular plants were found in the open water areas or loi.
Macroalgae –
One macroalga was found in loi and none was observed in the open water ponds. We
determined that the alga was Spirogyra however for further identification a sample was sent to
A. Sherwood, Botany Department, UH-Manoa. The alga has been included in Dr. Sherwood’s
NSF funded research to describe the diversity of freshwater algae in the Hawaiian Islands.
Fish –
There were 5 fish species found in the open water ponds and loi. All are considered introduced
species in Hawaii.
Objective #2: Trace biological responses before and during the shifts from an invasive species
dominated system to one under restoration.
Results indicate that there has been a change in open water area in the freshwater ponds, with a
loss of open water -49% between 2005 and 2010 (Figure 1). Comparisons to older satellite
photographs indicate that there may be periods of over growth and recovery so this loss may not
be a progression.
Figure 1. Changes in open water area at Heeia.
The four species of emergent vegetation the form canopies responded differently to a flash flood
event (4-5 June 2011). For example, in the largest pond on the property extensive areas of the
California grass canopy were torn away from the shoreline, swept downstream and were deposited
as three islands in the center of the pond (Figure 2a & b). In contrast, the aka akai canopy was not
uprooted, however a large floating mat of California grass was deposited in front of one of the aka
akai canopies.
Figure 2. a. Vegetation along margin of pond in February 2011 pre-storm. b. Vegetation along
margin of pond in June 2011 post-storm.
Objective #3: Record chemical and physical responses before and during the shifts from an
invasive species dominated system to one under restoration.
Coastal ponds
Dissolved oxygen (DO) saturation was ca. 40% in the coastal pond at the control location and
showed little diurnal change (Figure 3). In contrast, DO saturation in the taro loi ranged from ca.
50% in the hours before sunrise to >100% during the day. In the California grass canopy DO was
similar to the control and both nitrate and ammonium were low (Figure 4). A depth profile showed
that total suspended solids was low between 10 – 50 cm however was ca. 1500 mg l-1 at 75 cm
and composed mostly of organic matter. The Job’s tears canopy also had low DO saturation
throughout a 24 h period (Figure 5). In Job’s tears ammonium was the highest measured in any of
the canopies >5 uM at 75 cm. Neke was characterized by extremely low DO <5% saturation,
nitrate and ammonium, and total suspended solids (Fig). The Aka akia canopy had a diurnal signal
in DO with ca 20% saturation at night and ca. 45% during the day. Nitrate was highest in this
canopy ca. 1.2 uM at 10 cm (Fig).
Figure 3. Dissolved oxygen saturation in open water pond and taro loi of a 24 hour period.
Figure 4. Dissolved oxygen saturation in a California grass canopy and an unvegetated control
location. Ammonium and nitrate (uM) concentrations along a depth gradient in California grass
canopy. Total suspended solid, organic fraction and inorganic fraction in California grass canopy.
Figure 5. Dissolved oxygen saturation in Job’s Tears grass canopy and an unvegetated control
location. Ammonium and nitrate (uM) concentrations along a depth gradient in Job’s Tears grass
canopy. Total suspended solid, organic fraction and inorganic fraction in Job’s Tears grass canopy.
Figure 6. Dissolved oxygen saturation in Neke canopy and an unvegetated control location.
Ammonium and nitrate (uM) concentrations along a depth gradient in Neke canopy. Total
suspended solid, organic fraction and inorganic fraction in Neke canopy.
Figure 7. Dissolved oxygen saturation in Aka akai canopy and an unvegetated control location.
Ammonium and nitrate (uM) concentrations along a depth gradient in Aka akai canopy. Total
suspended solid, organic fraction and inorganic fraction in Aka akai canopy.
Taro loi.
Nitrate and ammonium concentrations associated with a flash flood event varied both spatially and
temporally (Figure 8). Pre-storm nitrate was ca. 2 uM at both stream sampling locations, during the
storm it increased to >3 uM and attenuated post storm to ca 1 uM 25 days later. The two sampling
locations in the loi reflected the spike in nitrate associated with the stream however because the
pre-storm nitrate was <0.5 uM in the loi, the change in nitrate concentration was much greater in
the loi than the stream. Also, the storm-associated spike in nitrate was only 2-3 days in the loi with
nitrate returning to pre-storm levels much faster in the loi than the stream. In contrast, ammonium
concentration did not change during the sampling period in the stream sites while a clear spike was
measured in the loi (Figure 9). After a 3 day post-storm lag ammonium concentration at the loi
enter sampling location increased from ca. 1 to >10 um and at the loi exit from 10 to > 40 um. At
both sampling location ammonium concentration dropped to below pre-storm levels.
Figure 8. Nitrate (uM) concentrations at four sites, pre- and post-storm.
Figure 9. Ammonium (uM) concentrations at four sites, pre- and post-storm.
Objective #4: Train interns and community volunteers to monitor and maintain their restoration
efforts.
Monitoring protocols for the four parameters were developed in tandem with the scientific
monitoring program. Detailed protocols for Kako‘o ‘Oiwi staff were created for each parameter,
which can also be used to train long-term interns in monitoring techniques. These are
accompanied by easy to follow guides with clear graphics and step-by-step flow charts that will
be used for short-term and single visit volunteer groups. Reliable and user friendly equipment
and supplies were employed to ensure that Kako‘o ‘Oiwi can continue the program well into the
future. The guides were printed and laminated in both handout and poster form for practical use
with small and larger groups at the Mahuahua ‘Ai o Hoi wetland site. In addition, laminated
wetland plants pocket guides and posters were also created.
Fig. 2 Examples of visual guides created for community based monitoring.
VII.
Applications
Outputs and management outcomes achieved. Outputs are defined as products (e.g.
publications, models) or activities that lead to outcomes (changes in user knowledge or
action). How did your project benefit resource managers? Has your project lead to
societal improvements?
A. Outputs
i.
New fundamental or applied knowledge
Objective #1: Collect baseline data by documenting the species occurring the He‘eia
inland fresh- and brackish water ponds and wetlands.
Species lists and vegetation maps produced have provided valuable baseline
data for future restoration efforts.
Objective #2: Trace biological responses before and during the shifts from an invasive
species dominated system to one under restoration.
Documenting how canopy forming species change during storm and non-storm events is
critical for managers’ ability to interpret the changes in vegetation at this site over varying
time periods. Identifying how these canopy-forming species respond to disturbance, such
as flash flooding, will inform restoration decisions. For example, we now understand that
California grass canopies readily fragment and that large mats of this species have the
potential take advantage of disturbance by redistributing in the open water and increasing
the coverage of this species. With the data we have provide in this study, managers at the
site can track the fate of the storm distributed vegetation mats of California grass. In
contrast, Aka akai canopies were fairly stable, even during storm disturbance but may be
subject to overgrowth by the storm distributed California grass mats. As a direct result of
this study managers now have clearer understanding of what it means to replace California
grass with Aka akai.
The loss of open water at the site has implications for waterbird and fish habitat also.
Implicating California grass in the loss of waterbird and fish habitat supports managers in
their efforts to control this invasive species.
Objective #3: Record chemical and physical responses before and during the shifts from
an invasive species dominated system to one under restoration.
Restoration of Hawaiian coastal wetlands has only been attempted at a few sites, with very
mixed results and little consideration for how that restoration affects downstream habitats.
More often than not pre-restoration data were not collected, and decisions were made with
little science-based understanding how to precede with restoration, hence the mixed
results. Also, communication between wetland managers and coral reef managers was
poor, at best. This project provides data that compares different vegetation types and their
affects on water quality. Our results also inform managers of trade-offs they will need to
make when one species is removed, and replaced with another. In other words, changes
in vegetation type impact water quality, and fish and waterbird habitat. Finally, the results
of this study were presented to coral reef managers and have a significantly improved and
refined their understanding of water quality in coastal freshwater systems that are linked to
marine habitats.
Objective #4: Train interns and community volunteers to monitor and maintain their
restoration efforts.
A. Outputs
i. A workshop was held during which Kako‘o ‘Oiwi outreach staff received training on
how to lead a monitoring session with volunteers for each parameter using the visual
guides that were created. Staff also
practiced the procedures with a group
of eight community research interns
that have been working with Dr.
Lemus for approximately 3 months
on another project. These interns
evaluated the guides and protocols
and gave feedback for their
improvement.
Community research interns
testing field protocols and guides.
ii. Informal presentations about the project and goals were given to community members
and visitors to the lo‘i. Presentations were also given at HCRI progress meetings.
ii.
iii.
iv.
v.
vi.
vii.
viii.
Scientific publications: None
Patents: none
New methods and technology: none
New or advanced tools (e.g. models, biomarkers): none
Workshops
A workshop presenting our results was held 31 August 2011 for the staff
of Kāko‘o ‘Ōiwi and TNC interns. Seven people attended.
Four water quality workshops for volunteers and staff of Kāko‘o ‘Ōiwi
were held April to June 2011 with a total of 23 people attending.
Presentations
Three presentations were given at the HCRI quarterly meetings.
Outreach activities/products (e.g. website, newsletter articles)
B. Management outcomes - I. Management application or adoption of:
i.
New fundamental or applied knowledge
We have learned that the invasive California grass is extending its cover over open
water wetlands and that the biology of this plant and of Neke creates regions of
anoxia. Water flowing through these plant canopies is very low in oxygen.
We are working with Kako‘o ‘Oiwi to develop plans for working to increase oxygen
levels throughout the open water regions of the wetland in the short term. In the long
term it is expected that lo’i will act as oxygenators of the water resources and
improve water quality.
i.
ii.
iii.
iv.
New or improved skills
We have trained a post-doc, student, and 4 community members in the use of
technical equipment (GPS; Water quality Sonde’s; ADVs).
Information from publications, workshops, presentations, outreach products
New or improved methods or technology
We are developing new techniques to link physical/chemical/ and biological
measures into and over all model of wetland function.
New or advanced tools
We are developing flow and nutrient models for different vegetation cover.
C. Management outcomes - II. Societal condition improved due to management action
resulting from output; examples:
i. Improved water quality - Controlling low oxygen levels with the He’eia wetland as planned
based on our data will increase water quality providing better habitat for native fish
and invertebrate and vertebrate animals.
ii. Reduced hypoxic zone area. This project directly addresses hypoxia in the wetland and
downstream areas. Our work has pointed out the role that California grass plays in
wetland anoxia and points to a need for eradication of California grass and
oxygenation of water in the wetland by opening up areas for taro loi.
D. Partnerships established with other federal, state, or local agencies, or other
research institutions (other than those already described in the original proposal).
We have developed new relationships with the Army Corp and the Hawaii
Department of Health.
VIII.
Evaluation
Describe the extent to which the project goals and objectives were attained. Provide
explanation for modification of goals and objectives.
Project Final Report
NOTICE
Subsequently, all NOAA COP recipients with approved grants will be asked to file a COP
Project Final Report in the specified format upon expiration or termination of grant support.
Consistency in reporting requirements for competitive research grant programs is desirable and
this is behind COP’s efforts in proposing a standardized format. The use of the Project Final
Report format will provide the level of detail required to evaluate the effort invested by
investigators and staff on project management; any actual accomplishments and research
findings; and what goals and objectives were attained. The proposed final report format is
compatible with the format in use by other agencies that participate in joint projects with COP,
e.g. the National Science Foundation.
Public reporting burden for this collection of information is estimated to average 600
minutes per response, including time for reviewing instructions, searching existing data
sources, gathering and maintaining the data needed and completing and reviewing the
collection of information.
Send comments regarding this burden estimate or any other aspects of this collection of
information, including suggestions for reducing this burden, to the National Ocean Service,
CSCOR/COP Office, 1315 East-West Highway, Silver Spring, MD 20910. Grant files are subject to
the Freedom of Information Act (FOIA). Confidentiality will not be maintained--the information
will be made available to the public. However, unpublished research results shall not be
published without prior permission from the recipient.
Notwithstanding any other provision of the law, no person is required to respond to, nor
shall any person be subject to a penalty for failure to comply with, a collection of information
subject to the requirements of the Paperwork Reduction Act, unless that collection of
information displays a currently valid OMB Control Number.