CHARACTERIZATION OF THE PHYSICAL ENVIRONMENT IN HE'EIA

FISHPOND, 'OAHU, HAWAI'I

A THESIS SUBMITTED TO

THE GLOBAL ENVIRONMENTAL SCIENCE

UNDERGRADUATE DIVISION IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

BACHELOR OF SCIENCE

GLOBAL ENVIRONMENTAL SCIENCE

MAY 2010

ByLindsey R. Benjamin

Thesis Advisor

Margaret McManus

I certify that I have read this thesis and that, in my opinion, it is satisfactory in scope andquality as a thesis for the degree of Bachelor of Science in Global Environmental Science.

THESIS ADVISOR

Margaret McManusDepartment of Oceanography

ACKNOWLEDGEMENTS

I would like to thank Margaret McManus, Kathleen Ruttenberg, Chip Young, Becky

Briggs, Jane Schoonmaker, and Rene Tada for their support during the writing of this

thesis.

This research was supported by a grant/cooperative agreement from the National Oceanic

and Atmospheric Administration; projects R/EL-42 and R/AQ-84, which are sponsored

by the University of Hawaii Sea Grant College Program, SOEST, under Institutional

Grants NA05OAR4171048 and NA09OAR4170060 from NOAA Office of Sea Grant,

Department of Commerce. The views expressed herein are those of the authors and do

not necessarily reflect the views of NOAA or any of its subagencies. UNIHI-

SEAGRANT-XB-10-01.

ABSTRACT

iv

The physical characteristics of Heeia Fishpond are important in studying the fishpond

system as a whole. Information about the flow rates and water exchange, the forces

influencing water movement, and the physical gradients in the pond is essential to

understanding the environment. Acoustic current meters were placed at the pond makahas,

or openings, to collect data to make rating curves. Pressure sensors were used to collect

pressure data for both flow rate calculations and frequency spectra. A YSI probe was

used at the makaha and internal sites to monitor surface and bottom temperature and

salinity. The majority �3 10! of the water exchange with the environment occurs through

the northeast corner, through sites TM and OM1. Most of the variability in the water flux

is due to the tides, though a shift in the winds from Trade to Kona does have some effect.

Surface and bottom gradients in temperature and salinity are due to colder, low salinity

river inputs in the northwest and warmer, high salinity inputs along the eastern side of the

pond. This thesis forms a physical framework for the continuing biological and chemical

studies in He'eia Fishpond.

TABLE OF CONTENTS

Acknowledgements.Abstract.

List of Tables.

List of Figures.Chapter 1: Introduction.

Hawai'ian Fishponds.He'eia Fishpond.Circulation Studies of Fishponds.

Chapter 2: MethodsStudy Site........................................................Instrumentation..

Data AnalysesWater Temperature and Salinity

Chapter 3: Results.Time-Series .

Rating Curves.Spectra.Flux

The Effects of Kona Winds on Water Flux

Temperature....................................................Salinity

Chapter 4: DiscussionChapter 5: Conclusions.References.

.1v

v1

.4 5

.7

.7

.8

12

14

15

15

20

28

31

36

.....37

38

40

43

42

~Pa e

2. Makaha tidal data for 03/10/08 to 04/10/08.

3. Rating curve formulas for all makaha.

.....36

Table

1. Instrument properties.

4. Kona wind flux change.....

LIST OF TABLES

.10

16

27

v1

v11LIST OF FIGURES

~Pa e

1. Map of Heeia Fishpond with study points marked..

2. Timeline of instrument deployment.

3. Depth for all makaha for 10/03/08-10/04/08 17-19

4. Rating curves for all makaha. 21-26

5. Spectra for all makaha. 29-30

6. Flux for spring flood, spring ebb, neap flood, and neap ebb tides.

7. Temperature at the surface and bottom on 11/17/07

32-35

37-38

8. Salinity at the surface and bottom on 11/17/07. .....39

CHAPTER 1: INTRODUCTION

Hawai'ian Fishponds

Ancient fishponds:

Ancient Hawai'i had a system of integrated agricultural practices that stretched from

mauka to makai mountain to ocean!. The land was divided into strips called ahupua'a

from inland to ocean, which were managed as a whole unit. The ali'i, or kings, allowed

the konohiki, or chiefs, to manage the ahupua'a; these konohiki divided their ahupua'a

into smaller sections to be managed by 'ohana, or individual families Costa-Pierce 1987,

Wyban 1992!.

Fishponds were a part of this system beginning as early as 1200AD Wyban 1992!, with

pond workers living on the ahupua'a under control of the konohiki. Fish were extensively

raised in the ponds, and the catch was distributed to everyone involved with the

ahupua'a: ali'i, konohiki, 'ohana, and other pond workers Costa-Pierce 1987!.

Four distinct types of fishponds existed in ancient Hawai'i: loko i'a kalo taro fishponds!,

loko wai freshwater fishponds!, loko pu'uone brackish water fishponds!, and loko

kuapa seawater fishponds! Costa-Pierce 1987, Kikuchi 1976!. Loko i'a kalo were

freshwater ponds that combined kalo or taro farming with raising of freshwater fish,

prawns, and green algae. Kalo was planted in rows of mounded earth, creating corridors

for fish to swim. Loko wai were much like loko i'a kalo without the kalo, although they

frequently had a small connection to the sea, making them technically brackish ponds

Kikuchi 1976!. Freshwater fish, prawns, and some fish that move from sea to freshwater

milkfish, mullet! were grown in ponds that had been carved out of the natural

topography. In addition, an edible mud was produced in loko wai Wyban 1992!. Ponds

that were situated on land with stream inputs but connected to the sea by a ditch were

loko pu'uone. A sluice gate or makaha was usually present in the ditch so inputs from the

sea could be shut off if desired, and the pond was kept separated from the sea by mounds

of sand Wyban 1992!. A wide variety of fish that could acclimate to both fresh and salt

water were kept in loko pu'uone.

The final type of fishpond, loko kuapa, was unique to Hawai'i, and a great achievement

in mariculture Costa-Pierce 1987!. The pond consisted of a walled-off section of a bay

or protected reef Wyban 1992!. Makaha in these ponds were made of a fairly fine mesh

of plant material that allowed the very young fish to pass, but not older fish; the smaller

fish from the sea would enter the fishpond to feed on the algae cultivated there, and could

only leave if they did not grow too large to fit back through the mesh. Thus, the pond was

stocked by nature instead of by man. The walls around these ponds ranged from 46m to

1920m in length, and were made of rocks and fill Costa-Pierce 1987!. Usually the walls

were left porous, but in a few ponds coralline algae was cultivated in the wall; coralline

algae secretes limestone which acts as mortar, sealing the wall Wyban 1992!. The walls

of loko kuapa were sloped, with a greater slope on the seaward side of the wall to

dissipate wave energy. Ponds ranged in size from 1 to 523 acres Wyban 1992!

Post-contact influence on fishponds:

In 1819, King Kamehameha II abolished the kapu system, or Hawai'ian system of laws

and religion Kikuchi 1976!. This system had, among other things, ensured the ali'i

complete control over their lands, resources, and people. It had functioned as a barrier

between invasive Western culture and native Hawai'ian culture, keeping the Hawai'ian

lifestyle alive Kikuchi 1976!. With its abolition came changes in Hawai'ian culture that

eventually lead to a near total Westernization of Hawai'ians.

Before 1848, all the land and resources in Hawai'i were property of the ali'i. However, in

a legislative move known as the Great Mahele ' division' !, land became a commodity

that could be bought and sold Costa-Pierce 1987, Wyban 1992!. With this decision, the

integrated agricultural system fell apart as ahupua'a were no longer kept whole, and

anyone could purchase land for other purposes. Without the support of the rest of the

ahupua'a, and with fishponds now viewed as an economic venture as opposed to a food

source, fishponds rapidly fell into disuse. In 1873 there were an estimated 241 ponds,

with 24 ponds covering 1046 acres in Kane'ohe Bay, 'Oahu. By 1975, there were only 28

ponds in Hawai'i, with 25 on 'Oahu; there are at least remnants of 12, with 7 in

functional condition Wyban 1992, Devaney 1982, Kelly 1976!. Most fishponds in

Kane'ohe Bay were filled shortly after World War II Devaney 1982!.

He'eia Fishpond

He'eia Fishpond is an 88 acre loko kuapa located in Kane'ohe Bay, 'Oahu, built about

300 years ago Kelly 1976!. It has one of the longest walls of all the fishponds �000 feet!

surrounding it completely, even on the landward side Devaney 1982, Kelly 1975!. The

seaward side of the wall is 12 to 14 feet thick and includes coral to induce the growth of

coralline algae, making it watertight Wyban 1992, Kelly 1975!. Because of this, the

walls had to be kept in good shape to keep the water level in the pond above the

surrounding bay Kelly 1975!. There were two very major flood events that damaged the

walls of He'eia fishpond. The first, in May 1927, destroyed a section of the North West

wall; part of this destroyed wall makes an island in the pond Kelly 1975!. The second

event, in May 1965, destroyed a 600 foot section of the South West wall. He'eia fishpond

is one of the last two working fishponds on 'Oahu Yang 2000!.

Currently, He'eia fishpond is managed by Paepae o He'iea POH!, a nonprofit group

devoted to restoring and caring for the pond. POH is currently removing invasive

mangroves and repairing the section of the wall damaged in 1965 using traditional

building methods. POH grows six species of fish and 5 species of crab, though these

animals are fed as opposed to relying on the pond environment to provide for them.

Circulation Studies in Hawai'ian Fishponds

There are very few studies of circulation patterns or water transport in Hawai'ian

fishponds. Ertekin et al. �996! modeled the circulation in One Ali'i fishpond on

Moloka'i using the finite element method to solve the conservation of mass and

momentum equations. The circulation was modeled with only one of the two makaha

open, as well as with both open. The water velocity through the makaha was higher in the

case with only one open makaha, but the water in the pond was circulated more when

both makaha were open. Ertekin et al. �996! concluded that both the number of makaha

as well as their location in relation to the forces at work were important in determining

and controlling the circulation in the fishpond.

Yang �000! modeled the effects of stream runoff and winds on circulation in fishponds.

The same modeling method was used as in Ertekin et al. �996!, but yielded only

qualitative results. The magnitude of the effect of stream runoff was found to depend on

the amount of runoff; a small amount changes the purely-tidal circulation pattern only in

the area into which the stream empties, and the water velocities in this area are very

nearly equal to the stream velocity. A larger amount of runoff changes a greater portion

of the circulation pattern, and at very high runoff it causes water to flow directly from the

stream mouth to the makaha and out of the pond. Wind was found to generate small

waves within the fishpond, but it did not have much of an effect on the purely-tidal

circulation pattern. Outside the pond, the wind had a much greater effect on circulation.

Wind was found to alter the rate of water flow through the makaha by altering the flow

just outside the makaha in the surrounding waters. The effects of wind on the circulation

are larger for larger fishponds.

CHAPTER 2: METHODS

Study site

He'eia fishpond is a walled section of Kane'ohe Bay on the windward side of Oahu

Figure I!. There are four ocean-side openings in the wall of the pond, designated Triple

Makaha TM!, Ocean Makaha 1 OM1!, Ocean Break OB!, and Ocean Makaha 2 OM2!.

There are three openings connecting the freshwater source RIVER! to the pond,

designated River Makaha 1 RM1!, River Makaha 2 RM2!, and River Makaha 3 RM3!

Figure I!. TM, OM1, OM2, and RM3 are inakaha, or gated openings. OB and RM2 are

depressions in the wall that water flows over, and RM1 has very diffuse flow.

21.43

21.43

21.43

21.43

21.43

21.43

21.43

21.43

157.811 157 81 157 809 157.808 157 807 157 806

Figure 1. He'eia Fishpond with makaha and stakes marked. The red circles on the

perimeter of the pond represent makahas, the red circles in the interior of the pond

represent water sampling locations stakes!.

Instrumentation

Two Nortek Aquadopp. Profilers were used to measure current speed and direction as a

function of water depth. Three acoustic beams measure a current profile for the water

column. Pressure measurements are made with a pressure transducer on the instrument.

Three Sontek Argonaut-SW. instruments were used to measure water depth as well as

flow speed and direction. A vertical acoustic beam determines depth from scattering at

the surface, while two other acoustic beams measure two-dimensional water velocity.

HOBO pressure data loggers were used to measure pressure.

The instruments were deployed in the pond in the timeframe indicated in Figure 2. At TM,

a Nortek instrument was deployed from 08/30/07 to 11/13/07 and from 11/15/07 to

02/10/08, and a Nortek with co-located pressure sensor were on the site from 02/13/08 to

05/08/08. A Sontek instrument was deployed at OM1 from 09/13/07 to 11/13/07,

11/15/07 to 02/10/08, and from 02/13/08 to 05/08/08. Another Sontek instrument was

deployed at OB for the same time periods. At OM2, RIVER, RM1, and RM2, pressure

sensors were in place from 08/08/07 to 11/29/07, 02/26/08 to 05/08/08, and 05/18/08 to

09/01/08. In addition, a Sontek instrument was at RIVER from 09/13/07 to 11/13/07,

12/06/07 to 02/10/08, and from 02/13/08 to 05/08/08. A Nortek was in place at RM3

from 08/30/07 to 11/13/07, 11/15/07 to 02/10/08, and 03/05/08 to 05/08/08.

Wind and tide measurements come from Coconut Island in Kane'ohe Bay.

10

Sample RateInstrument Feature

current speed

current direction

every 15 minutesNortek

current speed

Sontek current direction every 15 minutes

depth

every 15 minutesHOBO pressure

temperaturecontinuouslyYSI

salinity

Table 1. Instrument type, measurement, and sample rate.

I l'L"LJL'I il.

!LILf

I L'

I 'U

C'rl

E

O

e

E

0

E

11

12Data Analyses

Calculation of flow rate:

Using the water speed and depth from either the Sontek or Nortek instruments along with

the makaha dimensions, flow rates in m /s were determined for each of the rectangular

makaha TM, OM1, OM2, RM3! from the equation:

P = wdv sin�!

where w is the makaha width, d is the water depth, v is the magnitude of the water

velocity, and 0 is the direction of the water flow, with 0 = 0 corresponding to flow

directly through the makaha. For OB, water flows over a flat wall, so the flow rate in m /s

was also determined by Equation I!, with w as the length of the wall instead of makaha

width, and only for d>0.86m the height of the wall!. For RIVER, water flows through an

irregularly-shaped bed. A profile of bed depth across the site was made, and the areal

aspect of Equation I!, wd, was calculated from the profile assuming water fills in the

deepest parts of the bed first. For RM2, the wall that water flows over is at an angle with

respect to the ground, so that the areal part of Equation I! was calculated in the same

way as RIVER.

Construction of rating curve: 13

Spectral analysis:

Spectral analysis is the evaluation of the frequency components of a time series signal to

determine patterns of cycles in the data. The depth data for each makaha or pressure data

converted to depth! were evaluated using Fourier transforms to make a spectrum of

power in terms of frequency.

Flux volume calculations:

In order to determine the importance of flow through each makaha through a tidal cycle,

flow volumes for spring flood, spring ebb, neap flood, and neap ebb tides were examined.

Pressure data from each part of a tidal cycle was converted to flow volume using the

rating curves, and the total flux through each makaha for spring flood tide, spring ebb

tide, neap flood tide, and neap ebb tide was determined.

A rating curve is a curve relating the discharge from a river or outlet to the stage or height

of the water at the outlet. The curve is usually given as discharge as a function of stage,

though stage is represented as the vertical axis and discharge as the horizontal axis.

To construct a rating curve, the flow rates calculated above were averaged for every

depth, giving average flow rate discharge! as a function of water depth stage!. Curves

were fitted to this data, forming rating curves.

Water Temperature and Salinity: 14

A YSI probe was used to create depth profiles at each of the sites during water sample

collection. The upper 20 cm and lower 20 cm temperature and salinity data were

averaged for each site to create surface and bottom averages. For each property, a

colorbar was designed to fit the data ranges necessary, and dots of the appropriate color

were plotted at the location of the sample site with a map of the fishpond superimposed

on top. An attempt was made to produce an interpolated colormap for each property;

however, the location of the data points made artifacts within the colormap which were

not present in the data.

CHAPTER 3: RESULTS 15

Time-Series

Wind

During the period between 03/10/08 and 04/10/08 winds were dominated by north east

trades 90'/o! Figure 3a!. Winds from the south west i.e. Kona winds! occurred 10'/o of

this period. Trade winds ranged from 0.8 to 9.4 m/s with an average wind speed of 11.0

m/s. Kona winds ranged from 0.2 to 1.7 m/s with an average wind speed of 0.7 m/s

Tide

The tides in this region are semi-diurnal and mixed. During the study period spring tides

occurred around 03/13/08, 03/27/08, and 04/10/08 and neap tide occurred around

03/20/08 and 04/03/08 Figure 3!. The average water depth in Kane'ohe Bay was 0.29 m,

with a maximal tidal range of 0.44 m. The average water depths and maximum tidal

ranges for each makaha are reported in Table 2. Across all sites, the average water depth

was 0.55 m and maximal tidal range was 0.41 m.

Table 2. Makaha tidal data for 03/10/08 to 04/10/08. 16

Makaha

Abbreviation m! Range m!

0.430.57TM

OM1 1.10 0.37

Ocean Break OB 0.490.89

Ocean Makaha 2 OM2 0.43 0.40

River 0.430.38RIVER

River Makaha 1 RM1 0.47 0.38

River Makaha 2 RM2 0.45 0.38

River Makaha 3 RM3 0.14 0.39

Makaha

Triple Makaha

Ocean Makaha 1

Average Water Depth Maximum Tidal

01

~ 05

- ~ 05

01

- ~ 15

-0 2

- ~ 25

-0 3

04/1 304/0603/3003/2303/1 603/09

15

�~ 5

003/09 ~ 4/1 3~ 4/06~ 3/30~ 3/23~ 3/1 6

15

05

0~ 3/09 04/1 3~ 4/06~ 3/30~ 3/23~ 3/1 6

15

05

~ 4/1 304/06~ 3/3003/2303/1 6~ 3/39

15

~ 5

003/09 ~ 4/1 304/06~ 3/30~ 3/23~ 3/1 6

15

05

~ 4/1 3~ 4/06~ 3/3003/33~ 3/1 6~ 3/09

15

~ 5

003/09 ~ 4/1 304/06~ 3/30~ 3/23~ 3/1 6

15

05

~ 4/1 3~ 4/06~ 3/30~ 3/3303/1 6~ 3/09

05

~ 309

003/09

Figure 3. Wind a!, predicted tide b!, and depth for TM c!, OM1 d!, OB e!, OM2 f!,

RIVER g!, RM1 h!, RM2 i!, and RM3 j! from 03/10/08 to 04/10/08.

Rating Curves 20

Rating curves for flow into and out of TM, OM1, OM2, OB, and RM3 were made.

RIVER and RM2 have only a single rating curve because of unidirectional flow. The

rating curves for TM in, TM out, OM1 in, OM1 out, OB in, OB out, and RIVER are

logarithmic curves of the form of:

y � b

�!x=ae '

where flow volume is x, water depth is y, a is the amplitude, b is the offset in the

horizontal asymptote, and c is the degree of curvature. The rating curves for OM2 in and

OM2 out are partial Gaussian curves of the form of:

- y-cj'

x = a+be �!

�!x = ay+ b

for flow volume x, depth y, slope a, and intercept b. The rating curve for RM3 in is an

asymmetric Gaussian curve formed by joining two partial Gaussian curves of the form

given in Equation �!. Actual formulas for the rating curves are given in Table 3.

where flow volume is x, depth is y, a is the horizontal displacement, b is the amplitude, c

is the offset in the peak, and d is the width of the bell. The rating curves for RM2 and

RM3 out are linear curves of the form:

09

0 ~

07

06

6 05

04

03

02

01

~ 2~ 1~ 000040 02 ~ 06discharge m3/sec

09

0 ~

07

06

6 05

04

03

02

01

~ 015~ 01~ 0050 ~ 02 0 025 ~ 03 ~ 035 ~ 04 ~ 045 005discharge m3/sec

22

15

~ 9 ~ 01 ~ 02 ~ 03 004 01

15

14

13

~ 9 002 ~ 04 ~ 08 0 08discharge m3/sec

~ 1 ~ 12 014

13E

12

F

12

~ 05 0 05 0 07 ~ 08 ~ 09discharge m3/ssc

23

12

6 m ~ 9

08

~ 7

7050201 ~ 30 40discharge m3/sec

13

12

E 5, 09

08

~ 7

~ 6

~ 5 605040201 ~ 30discharge m3/sec

24

07

~ 65

06

~ 55

05E

~ 45

04

~ 35

03

~ 25

02~ 35~ 30 2501~ 05 015 02

discharge m3/sec

07

E 05

04

03

020 ~ 2502~ 05 ~ 1 015

discharge m3/sec

09

085

0 ~

075

Ere 07

065

055

0505~ 40201 ~ 3

discharge m3/sec

~ 61

0 605

~ 6

60 595

~ 59

0 585

~ 58 ~ 45~ 4~ 350302~ 15~ 1 ~ 25discharge m3/sec

~ 9

0 85

~ ~

0 756

~ 7

0 65

~ 6

0 55 ~ 09~ 08~ 07~ 06~ 03~ 02~ ~ 1 004 ~ 05discharge m3/sec

26

0 55 0 ~ 04discharge m3!sec

Figure 4. Rating curves for TM in a!, TM out b!, OM1 in c!, OM1 out d!, OB in e!,

OB out f!, OM2 in g!, OM2 out h!, RIVER i!, RM2 j!, RM3 in k!, and RM3 out I!.

27

Site Direction Formula

x = 0.9+ 0.181og0.15

TM

x = 1.02+ 0.126 log0.18

Out

x =1.43+ 0.081og0.15

OM1

x = 1.38+ 0.061og0.005

Out

x = 0.9+ 0.07 log y!

OB

x = 0.44+ 0.08 log0.01

Out

y-o.vl!'

x = 0.75 + 0.18e

OM2

y-o.vs!'

x = 0.017+ 0.15eOut

x = 0.86+ 0.07 log0.34

RIVER Both

x = 0.06x+ 0.5755RM2 Both

y � 0.726!'

x = 0.03+ 0.058e '"' for x 0.726

y-o.vzo!'

x = 0.009+ 0.052e '"" for x ! 0.726RM3

x = 0.8 � 3.85yOut

Table 3. Formulas for rating curves given in Figure 4.

Spectra 28

The spectra show the frequency dependence of the water depth data Figure 5!. The

spectrum for TM shows strong signals at 12 and 24 hours, with a larger and sharper peak

at -0.5 days. The OM1 power spectrum has strong 12 and 24 hour signals, with the 24

hour signal having a larger spread. The spectrum for OB has a strong, slightly spread 24

hour signal with a weaker 12 hour signal. Strong 12 and 24 hour signals are present in the

OM2 spectrum, with the 12 hour signal having a sharper peak with a slightly larger

amplitude. The spectrum for RIVER shows strong, nearly equal signals at 12 and 24

hours, with the 24 hour peak having a larger spread. The RM1 power spectrum has strong

12 and 24 hour signals, with the 24 hour signal having a larger spread. The spectrum for

RM2 has a strong, slightly spread 24 hour signal with a strong, sharply peaked 12 hour

signal. Strong 12 and 24 hour signals are present in the RM3 spectrum, with the 12 hour

signal having a sharper peak with a slightly larger amplitude.

29

~ ~

Ere ~CLCL

W ~ 4CL

~ 2

00 105Penod Days!

~ ~

XE ~ 6CL

3 ~ 4

~ 2

100 5Penod Days!

E e QEo

3 04o

02

0 105Period� Days!

XE o Qoo

s 04o

02

00 105Period Days!

30

~ ~

E ~ 6CL

s 04CL

~ 2

0 4 5 6Penod Days!

~ ~

E 06CL

s ~ 4CL

~ 2

00 5Penod Days!

~ ~

Ee ~

s 04

~ 2

105Pened Days!

~ ~

6 ~ 6CL

a ~ 4CL

~ 2

00 105Pened Days!

Figure 5. Spectra for TM a!, OM1 b!, OB c!, OM2 d!, RIVER e!, RM1 fj, RM2 g!,

and RM3 h! from data from 03/10/08 to 04/10/08.

31Water Flux

The water flux through the makaha during spring flood tide Figure 5a! shows most water

passing through OM1 �.4 10 of the total flux during spring tide and neap tide!, RIVER

�.5 10!, and TM �.0 10!. A lower volume of water flows through OB �.6 10!, and a

negligible amount of flow moves through RM2, OM2, and RM3 channel �.5 to, 0.4 to,

and 0.2 10, respectively!. All makaha have water entering the pond with the exception of

RM3 during spring flood tide.

During spring ebb tide Figure 5b!, the greatest volume of water moves through OM1,

TM, and RIVER �2.5 10, 8.8 10, and 4.8 10, respectively!, while a negligible amount of

water moves through RM2, OB, OM2, and RM3 �.3 to, 1.3 to, 0.4 to, and 0.2 10,

respectively!. During spring ebb, all ocean-side makaha have water leaving the pond,

while all river-related makaha allow water to enter the pond.

During neap ebb tide Figure 5d!, the greatest volume of water flows through OM1, TM,

and RIVER �1.8 10, 8.3 10, and 7.1 /0, respectively!. A moderate to negligible amount of

During neap flood tide Figure 5c!, the greatest volume of water flows through OM1 and

TM �.8 10 and 4.8 10, respectively!. A moderate volume of water flows through OB, RM3,

and RIVER �.0 to, 2.5 to, and 2.4 10, respectively!, and a negligible amount of water

flows through RM2 and OM2 �.0 10 and 0.4 10, respectively!. During neap flood tide, all

makaha have water entering the pond with the exception of RM3.

32

water flows through OB, RM2, OM1, and RM3 �.2 to, 1.3 to, 0.3 to and 0.2',

respectively!. During neap ebb tide, all ocean-side rnakaha have water leaving the pond,

while the others have water entering the pond.

21 43

21 43

21 43

21 43

21 43

21 43

21 43

21 43

157 811 157 81 157.889 157 808 157 807 157 806

33

21. 43

21 43

21 43

21 43

21 43

21 43

21 43

21 43

157 811 157 81 157.809 157 808 157 807 157 806

34

21. 43

21 43

21 43

21 43

21 43

21 43

21 43

21 43

157 811 157 81 157.809 157 808 157 807 157 806

35

21. 43

21 43

21 43

21 43

2'I 43

21 43

21 43

21 43

157 811 157 81 157.809 157 808 157 807 157 806

Figure 6. Water flux as a percentage of total spring and neap water flux for spring flood

a!, spring ebb b!, neap flood c!, and neap ebb d!.

The Effects of Kona Winds on Water Flux 36

Kona Wind Flow % Change from Trade Wind Flow!Site

Flood Ebb

-43.9 13.0TM

-44.0OM1 17.3

-36.5OB 75.4

-33.8OM2 24.2

-24.2 22.7RIVER

20.1 20.1

-11.4RM3 409.6

During spring flood tide with Kona winds, flow through all sites was below normal

except RM2 and RM3, which were significantly increased. During spring ebb tide with

Kona winds, flow through all sites except RM3 was above normal.

As previously mentioned, trade winds dominated for the majority of the study 90%!.

Kona winds were present during 10% of the study. The volume of water moving through

each makaha was calculated for spring flood and spring ebb tides under Kona wind

conditions Table 4!.

Table 4. Kona wind and trade wind water flux comparison.

37Temperature

21 439

2621 438

25 75

25 521 437 25 25

25

24 7521 436

24 5

24 2521 435

24

23 75

21 434 23 5

23 25

2321 433

21 432

157 81 157 809 157 808 157 807 157 806

During water sampling on 11/17/07, water temperatures in the pond ranged from 23 to 26

degrees C Figure 7 a and b!. The surface temperatures are coldest at RM2 and OM2 and

warmest at S 1 and S3. The bottom temperatures are coldest at RM2 and OM2 and

warmest at Sl, S3, and OCN1. At S6, S8, S9, S13, OCN1, and OB, surface temperatures

are cooler than bottom temperatures.

38

21 439

2621 438 25. 75

25 5

21 437 25 25

25

21 436 24 75

24 5

24 2521 435 2423 75

21. 434 23 5

23 25

2321 433

21 432

157.81 157 809 157 808 157 807 157 806

Figure 7. Temperature on 11/17/07 at the surface a! and bottom b!.

Salinity

During sampling on 11/17/07, the salinity of the pond ranged from 0 to 33 PSU Figure 8

a and b!. At the surface, the salinity is lowest at RM1 and RM2 and highest near the

south-eastern corner of the pond. At the bottom, the salinity is lowest lowest at RM1 and

RM2 and highest near the south-eastern corner of the pond. Salinity at RM1 is

significantly higher at the bottom than at the surface. At both surface and bottom there is

a clear gradient between high and low salinities.

39

21 439

3321 438 3027

21 437 24

21

21 436 18

15

1221 435

21 434

21 433

21 432

157 81 157 809 157 808 157 807 157 806

21 439

3321 438 30

2721 437

24

21 436 18

15

1221 435

21 434

21 433

21 432

157 81 157 809 157 808 157 807 157 806

Figure 8. Salinity on 11/17/07 at the surface a! and bottom b!.

CHAPTER 4: DISCUSSION 40

In terms of flux during normal conditions, OM1 and TM allow the most water transport,

making the north east corner of the pond the most influential inlet/outlet area. The

RIVER site also shows fairly large flow volumes, particularly during neap ebb. Flux

through the north east corner of He'eia fishpond is greater during ebb tides than during

flood tides because of input from the river makaha; while water may back up during

flood tides to the river makaha, river flow is downstream and eventually into the pond.

When comparing the flux during normal conditions and the flux during Kona wind

conditions, interesting patterns emerge Table 3!. During flood tide, the trade wind fluxes

are all larger except for RM2 and RM3. During ebb tide, the fluxes are all smaller with

The spectra all have peaks at roughly 12 and 24 hours, thus the tidal forces appear to be

the most important force in the cyclical exchange of water between the pond and its water

sources/sinks. The sharpness of a spectral peak is indicative of the accuracy in timing.

That the 12 hour peaks are sharper than the 24 hour peaks could represent that the 12

hour cycle is more precise in its timing than the 24 hour cycle. In all the spectra except

OB, the 12 hour peak is larger in magnitude than the 24 hour peak. This anomaly is likely

again! due to the fact that OB is a wall instead of makaha, so it does not allow free

exchange at all times. Interestingly, even RIVER, RM1, RM2, and RM3 also show tidal

influences; this could be because the tide runs up the river and influences the river

makahas.

trade winds than with Kona winds. There does appear to be an influence on wind 41

direction on the flux through the makahas of the fishpond. Keeping in mind that south-

west Kona winds blow in the opposite direction as the north east trades winds.

Trade winds aid in the transport of water through makahas and into the fishpond during

spring tide. While Kona winds aid in the transport of water out of the pond during ebb

tides.

An obvious northwest-southeast temperature gradient exists on both the pond surface and

bottom. The temperatures in the northwest are colder than those in the south east. This

gradient is caused by the inflow of colder water from the river makaha and warmer water

from the ocean makaha.

There is a salinity gradient from the northwest to the eastern side of the pond. The

salinity of the river water is essentially 0, while the salinity of the ocean water is 33 PSU.

The gradient is from the river water input in the North West corner to the ocean water

inputs from the makaha TM, OM1, OB, and OM2! on the eastern side of the pond.

CHAPTER 5: CONCLUSIONS 42

The majority of the flux into and out of the pond �3 10! occurs at the northeast corner,

through TM and OM1. Inputs from the river through RM2 and RM3 make up only 7 10.

The other 30 10 flows over OB, or through OM 1 and the RIVER.

The winds force water movement into and out of the fishpond. Kona winds, which are a

reversal of the normal northeast trade winds, cause a -44 10 decrease in flow through TM

and OM1 during spring flood tide, when winds push against the water flowing into the

pond. Kona winds also cause a -15 10 increase in flow through TM and OM1 during

spring ebb tide, when winds push with the water flowing out of the pond.

There are northwest-southeast gradients in temperature and salinity in both the surface

and bottom waters of He'eia fishpond. Cold freshwater inputs from the river enter at the

northwest end of the pond, while warm saltwater enters from Kane'ohe Bay along the

eastern side of the pond.

The tides are a major force acting on He'eia fishpond, as evidenced by the strong -12 and

-24 hour signals in the frequency spectra that match the Hawai'ian mixed diurnal tides.

The spectra for all makaha show these strong signals, indicating that tides affect water

flowing through all makaha, even the river makahas.

43REFERENCES

Devaney, D.M. Kaneohe: A History of Change. Bess Press. 1983. 271p.

Ertekin, R. C., H. Sundararaghavan, and A. T. F. M. Van Stiphout �996!, Molokai

fishpond tidal circulation study, Technical Report OE Report No.: UHMOE�

96203, Sea Grant Report No.:UNIHI-SEA GRANT-TR-96-03, Dept. of Ocean

Engineering, University of Hawaii at Manoa.

Kelly, M. Heeia Fishpond, a Testament to Hawaiian Fish-Farming Technology.

Department of Anthropology document prepared for Pauahi Bishop Estate. Bernice

Pauahi Bishop Museum, Honolulu, Hawaii. 1976. 21p.

Kelly, M. Loko I'a o He'eia: Heeia Fishpond. Department of Anthropology document

prepared for Pauahi Bishop Estate. Bernice Pauahi Bishop Museum, Honolulu, Hawaii.

September 1975. 56p.

Kikuchi, W.K. �976!. Prehistoric Hawaiian Fishponds. Science 193 �250!. p 295-299.

Wyban, C.A. Tide and Current: Fishponds of Hawaii. University of Hawaii Press. 1992.

192p.

Costa-Pierce, B.A. �987!. Aquaculture in Ancient Hawaii. Bioscience 37 �!. p 320-331.

44

Yang, L. �000!. A circulation study of Hawaiian fishponds. Master thesis in ocean

engineering. University of Hawaii at Manoa.