Amadeo S. Timbol
Michael H. Kido
Donald E. Heacock

Prepared for

Limahuli Garden and Preserve (National Tropical Botanical Garden)
December 1989


Limahuli Stream drains a watershed (Limahuli Valley) of about 1.6 square miles (4.1 square kilometers) on northern Kaua’i Island (see map). It's headwaters arise from nearly 3,000 foot elevation, accretes, and flows 8.5 miles (13.7 km) before emptying into the ocean at 22o 13' 41" N, 159 34' 37" W.

Limahuli valley may be divided into lower and upper parts. Lower Limahuli consists of the drainage area up to the waterfall (1,560 ft elevation), while the upper portion is comprised of nearly pristine forest in an inaccessible ‘hanging valley’ above the waterfall. At present, a portion of lower Limahuli is utilized as a "Garden Area" by Limahuli Garden (National Tropical Botanical Garden). This study only involves that portion of the stream that flows through lower Limahuli Valley.

Scope and Objectives

The main purpose of this study was to compile a short-term baseline biological description of lower Limahuli Stream. This report covers two field work days on July 29 and September 2, 1989.

The objectives of the study were to:

1). Compile an aquatic macrofaunal list of both scientific and local or common names.
2). Make a semi-quantitative population estimate of fish, decapod crustaceans and stream macrobenthos in Limahuli Stream.
3). Describe the stream's physicochemical characteristics, i.e. dissolved oxygen, water temperature, pH, and conductance.
4). Describe the stream channel for width, depth, flow velocity and stream substrate using the designated sampling stations as representative habitat.
5). Identify riparian vegetation on both banks of the sampling stations and estimate the vegetative canopy covering the stream channel.


Sampling Stations

Four sampling stations were selected studied:

Station I- referred to as "Limahuli Falls" - 800 ft (243.8 km) elevation
Station II- "Stream crossing" - 550 ft (152.4 km) elevation
Station III- "Road crossing" - 200 ft (61.0 km) elevation
Station IV- "Above estuary" - 40 ft (12.2 km) elevation

Methods for Sampling Biota

Fish and crustaceans

Fish and crustaceans were collected with the use of a Coffelt BP-6 gas electroshocker. In electroshocking, a 3-ft scissor net was set downstream to collect those missed by the electroshocker operator. All specimens were collected from an estimated area of 20 x 1 m whenever possible. On the occasion where this was not possible, the biological count was extrapolated accordingly. Generally, electroshocking was continued beyond the 20 m zone to determine if additional species were present. If new species were found, they were added to the species list. Animals collected were identified, counted, and released live in the same general area.

Aquatic macrobenthic animals

The benthic invertebrate population was sampled with a WIDCO Surber Stream Bottom Sampler which had a frame of 30 cm x 30 cm (12" x 12") and a net mesh size of 728 microns. The procedure required the seating of the lower frame of the sampler on the stream bottom and agitating the substrate inside the frame so that bottom-dwelling animals drifted into the net attached to the frame. The contents of the net were stored in 70 % alcohol in plastic whirl-paks and brought to the laboratory for sorting and identification. Invertebrates were counted and identified to the lowest possible taxonomic level. Samples were generally taken in riffles with cobble bottoms at between 30 and 60 cm (12" and 24") water depth.

Riparian vegetation

For the purposes of this study, an estimate of relative cover and abundance (Table 1) for major plant species observed in the vegetative zone adjacent to the stream (riparian habitat) was adapted from Elliot and Hall (1977). Most scientific names are from Neal (1965).

Table 1. Vegetative coding system used in categorizing riparian vegetation (from Elliot and Hall 1977).

Cover Code Abundance Code
< 5% 1 rare (1- 4 plants) R
5% - 25% 2 occasional (5 - 14 plants) O
26% - 50% 3 frequent (15 - 29 plants) F
51% - 75% 4 abundant (30 - 99 plants) A
76% - 100% 5 very abundant (> 100 plants) V

The riparian area sampled was confined horizontally (across the stream) to within 3 meters of the waters edge and vertically (upstream-downstream) to within 10 meters on either side of the center of the sample station. Vegetation beyond this area was noted only if the canopy of the stream was involved or if interesting native plants were observed.

As significant physical differences may exist between the right and left banks of a stream, each was treated separately (note: left and right was always determined looking upstream). A riparian vegetation list was therefore compiled separately for each bank and generally plants are listed beginning with plants closest to the waters' edge. An asterisk (*) after a species indicates a tree species involved in creating the canopy over the stream channel.

The vegetative cover of the stream (canopy) affects the amount of sunlight reaching its surface. This may be another important factor in the over-all energy production of the ecosystem. The canopy was estimated as % shaded. A 100% shaded condition indicates that the canopy completely covered the stream channel, 50% shaded indicates that the canopy covered half the stream channel, and so on.


Sampling Physicochemical Features

Channel width, depth, and flow velocity

Channel width was obtained by stretching a 100-ft tape across the channel. Depth was measured from one bank to the other at one foot intervals at the upper two stations and at three-foot intervals at the lower two stations using the calibrated Swoffer flow meter for depth. Flow values obtained are from 0.6 of depth. The flow meter is accurate to within 1% and precision is a standard deviation of plus or minus 0.01 ft/s.


The substrate of each of the sampling stations was determined by above water visual examination. The substrate was sketched, photographed and quantified using a modified system adapted from the Wentworth classification of particle size (Table 2). The plant cover of exposed and submerged substrate was also estimated as it may be an important habitat component affecting species composition and abundance.

Table 2. Substrate coding system (modified Wentworth: Bovee and Cochnauer 1977) used in characterizing the substrate.

Substrate Particle size, range (mm)
1. Bedrock solid, lava slab
2. Exposed boulder 250 - 4,000 mm (10 in. - 13 ft)
3. Submerged boulder same as above
4. Cobble 65 - 250 mm (2.6 in. - 10 in.)
5. Gravel 5 - 65 mm (0.2 in. - 2.6 in.)
6. Sand 1 - 5 mm (0.04 - 0.2 in.)
7. Silt 1 mm (0.04 in.)
8. Plant detritus/organic material established percentage covering sampling area

Water Temperature

An alcohol thermometer was used for these data. Water temperature was cross checked with the oxygen meter and conductivity meter. These meters are also equipped to measure water temperature.


Water conductivity was measured with a YSI model 33 meter at subsurface in the same place where dissolved oxygen was measured. The meter has an accuracy of plus or minus 2.5% maximum error. Conductivity is expressed in micromohs/centimeter (umhos/cm).

Dissolved oxygen

This was measured with a YSI 57 dissolved oxygen meter at subsurface from an area representative of the sampling station. The meter measured oxygen in mg/L. The data were converted to per cent saturation. The meter accuracy is given at 0.1 mg/L.


This feature was measured with a Digisense pH meter model 5994 (Cole-Parmer Instrument Co.) at subsurface level at the same place where dissolved oxygen and conductance were measured. The accuracy for this meter is 0.01 pH unit. The meter was calibrated at each sampling site in accordance with the procedure manual.


Physicochemical Features

Limahuli Stream is small if compared with nearby Wainiha or Hanalei rivers. Its substrate consists of lava bedrock and large boulders at high elevations and boulder, cobble and gravel at low elevations. Its water is cold, has low conductance, is slightly acidic to slightly basic, and is oxygen saturated.

Channel width, depth, and flow velocity

Limahuli stream is characterized by narrow stream channels in the upper elevations, from 5 to 10 ft width, widens at mid elevations (15 ft) and is widest at the lowest elevation. It is shallow, only 0.4 ft at the upper elevations, becoming a foot deep at mid elevations and a little bit more than two feet at the lowest elevation. Its flow velocity is faster at the upper elevation (1.8 ft/s), slowing to a third of its original velocity at mid elevations, down to one-sixth at the lowest elevation. The data obtained is summarized in the table that follows (Table 7):

Table 3. Width, depth and flow velocity in Limahuli stream, Kauai. (July - September 1989).

Stations Parameters

Width (ft) Depth (ft) Flow Velocity (ft/s)
I 5 0.4 1.8
II 10 1.1 0.8
III 15 1.5 0.8
IV 30 2.2 0.3


The substratum is an important physical parameter of lotic ecosystems. In Hawaii, the substrate is characterized by solid lava bedrock, varying sizes of exposed or submerged weathered basalt particles, plant detritus, organic material, and varying degrees of vegetation covering the substratum. Substrate composition has important biological implications because it determines available aquatic habitat and affects physicochemical parameters such as dissolved oxygen.

Slope is an important determinant of substrate type as it affects the velocities of stream flows and the resultant scarification of the streambed. Limahuli descends from an elevation of approximately 1590 feet at the falls to sea level in less than 2.5 miles. High velocities are evidenced at higher elevations in the narrowness of the channel and the predominance of bedrock. High channel velocities may also be a factor in the low abundance of algae observed on the substrate throughout the stream system.

As in any lotic system in Hawaii it should be noted that substrate parameters, especially in the lower portions of the stream where velocities are amplified, are highly variable and subject to drastic change during periods of flood.

Table 4. Substrate in the four sampling stations at Limahuli stream, Kauai. (Wentworth scale: Bovee and Cochnauer (1977).

Station I: Limahuli Falls

Substrate % Coverage Estimated Plant Cover %
1/ bedrock 80 3
2/ exposed boulder 10 3
3/ submerged boulder 5 3
4/ cobble 3 <1
5/ gravel <1 -
6/ sand <1 -
7/ silt <1 -
8/ organics/detritus not estimated -

Station II: Stream crossing

1/ bedrock 40 5
2/ exposed boulder 10 5
3/ submerged boulder 25 5
4/ cobble 15 2
5/ gravel 5 <1
6/ sand 4 -
7/ silt 1 -
8/ organics/detritus not estimated -

Station III: Road crossing

1/ bedrock 5 <1
2/ exposed boulder 5 5
3/ submerged boulder 25 5
4/ cobble 30 5
5/ gravel 20 <1
6/ sand 10 <1
7/ silt 5 -
8/ organics/detritus not estimated -

Station IV: Above estuary

1/ bedrock 1 -
2/ exposed boulder 35 <5
3/ submerged boulder 30 <5
4/ cobble 25 <2
5/ gravel 5 -
6/ sand 2 -
7/ silt 2 -
8/ organics/detritus not estimated -

Temperature, conductance, dissolved oxygen, pH

These short-term data reflect only the conditions at the time of sampling and may be useful for comparative purposes only. On the other hand, Limahuli stream has not been studied before as far as the above parameters are concerned. These data are summarized in Table5.

Subsurface water temperature was determined to be between 18 and 19 degrees C with the uppermost Station I (Limahuli falls) only one degree cooler than the lower stations. This range is well within the range of unaltered streams in Hawaii (Timbol and Maciolek 1978) and within the living (tolerance) limit of native gobies (Hathaway 1978).

Conductance indicates total dissolved solids in water (Cole 1979). Results show a very low conductance, ranging from a low 59 microohms (umhos) at the upper Station I, gradually decreasing with decreasing elevation: to 61 at Station II, 68 at Station III, and to 71 at the lowest Station IV. These values are much lower than those for Kauai streams which have farms within their drainage areas (ave. 131 umhos, Timbol and Maciolek 1978). On the other hand, conductance for Limahuli stream compares favorably with those obtained at Wainiha River, a nearby but larger stream (53 through 83 umhos, Timbol 1986).

The pH was from slightly acidic at the upper stations to slightly basic at the lowest station. This is well within the range of neutral to slightly alkaline water for normal Hawaiian streams. This condition (acidic at the upper elevations, basic at the lower elevations) comes from organic acids of decomposing plant materials in the watershed which tend to make stream water acidic, but is neutralized as the water flows over volcanic rock.

Dissolved oxygen was measured as mg/L and converted to percent saturation. Limahuli stream water is oxygen saturated, from the upper elevations down to the lowest elevation. Considering the low water temperature, the oxygen available for the aquatic animals is considerable, from 9.26 to 9.45 mg/L (actual values obtained). These indicate clean, high velocity, bubbling waters.

Table 5. Summary of water temperature, conductance, dissolved oxygen, and pH obtained in Limahuli stream, Kauai (July - September 1989).


Sampling Stations

Water temperature
(degree Celsius)
18 19 19 19
Conductance (umhos/cm) 59 61 68 71
pH 6.20 6.63 6.77 7.44
Dissolved oxygen
(per cent saturation)
99 102 104 102

Riparian Habitat

Riparian vegetation (i.e. vegetation along side a stream or river) may be an important source of energy input for lotic (flowing-water) ecosystems. Allochthonous material has been shown to play a significant role in energy input for woodland streams in temperate climates (Minshall 1967, Fisher and Likens 1973). However, whether or not this is also true for Hawaiian streams has yet to be studied.

The stream channel appears to be highly colonized by non-native plants as evidenced by the common presence of yellow guava (Psidium guajava) and weedy species like honohono (Commelina diffusa) (Table 6). Clidemia hirta, considered an obnoxious weed by many, was found at station II and also elsewhere in isolated pockets.

The common presence of taro (Colocasia esculenta) along the water's edge indicates its cultivation in previous times. Significant remains of walled terraces (lo`i) alongside the stream may indicate the use of the stream by organized agricultural communities in ancient times.

Perhaps the most significant impact of alien plants on the stream itself was found in the extensive closed canopy created by tree species. The channel is shaded throughout most of its length, significantly reducing sunlight reaching the water's surface. Yellow guava (Psidium guajava) was the most common component of the riparian canopy although the octopus tree (Schefflera actinophylla) and the autograph tree (Clusia rosea) also contribute significantly to the canopy in lower portions of the stream.

In one fairly long section of stream above station II, hau (Hibiscus tiliaceus) has completely overgrown the channel making passage very difficult. The dense root mats in the stream created by such a condition may block the upstream migration of amphidromous species like the goby fishes (`o`opu). A large, very dense stand of another introduced species, wild bamboo (Schizostachyum glaucifolium), located on high ground above station II also contributes to the considerable shading of a significant portion of the stream channel.

Despite the common presence of introduced plants, significant patches of native vegetation still exist. The common native streamside plant mamaki (Pipturus albidus) for example, was found at station I although in a less dominant condition than observed in other streams. Also on the walls of the canyon were a few individuals of endemic Gahnia kauaiensis. The native fiber plant olona (Touchardia latifolia) was observed on the trail as was the Kauai endemic white hibiscus (Hibiscus waimeae subsp. hannerae).

It should be mentioned that streamside vegetation is highly susceptible to high water during flood. It is not uncommon therefore to see the vegetation laid down and strands of loose vegetation hanging incredibly high on the banks. Riparian vegetation is therefore constantly in a state of flux and only very hardy species can survive for any length of time alongside the stream. This may be one factor which clears the way for hardy species like yellow guava (Psidium guajava) and other introduced tree-like plants which now dominate the water's edge.

Table 6. Riparian vegetation in sampling stations at Limahuli stream, Kauai (July-September 1989).


Canopy- 90% shaded

Coverage Abundance

Left bank

taro (Colocasia esculenta) 1 R

small club moss (Selaginella arbuscula) 2 A

`ama`u fern (Sadleria sp.) 4 A

mamaki (Pipturus albidus) 4 A

yellow guava * (Psidium guajava) 4 A

Right bank

maidenhair fern (Adiantum raddianum) 2 0

juvenile `ama`u fern (Sadleria sp.) 5 V

kamole (Polygonum sp.) 5 V

Cyperus odoratus 2 R

yellow guava * (Psidium guajava) 5 V

ti leaf (Cordyline terminalis) 2 0

`ie`ie (Freycinetia arborea) 1 R

Gahnia kauaiensis 1 R



Canopy - 60% shaded

Coverage Abundance

Left bank

laua`e fern (Phymatosorus scolopendria) 4 V

kamole (Polygonum sp.) 5 V

honohono (Commelina diffusa) 5 V

air plant (Bryophyllum pinnatum) 1 R

wild ginger (Zingiber zerumbet) 5 V

Clidemia hirta 1 R

ti (Cordyline terminalis) 2 0

yellow guava * (Psidium guajava) 5 F

octopus tree (Schefflera actinophylla) 2 R

Right bank

Cyperus odoratus 4 0

Jamaica vervain (Stachytarpheta

jamaicensis) 1 R

wild ginger (Zingiber zerumbet) 2 F

yellow guava * (Psidium guajava) 5 F

kukui * (Aleurites moluccana) 3 0

octopus tree (Schefflera actinophylla) 1 R

`ohi`a-lehua (Metrosideros polymorpha) 1 R



Canopy- 100% shaded

Coverage Abundance

Left bank

Cyperus odoratus 2 0

honohono (Commelina diffusa) 3 F

Wedelia trilobata 5 V

hala (Pandanus tectorius) 3 F

yellow guava * (Psidium guajava) 3 F

autograph tree * (Clusia rosea) 3 F

Java plum * (Syzygium cumini) 3 0

Right bank

Wedelia trilobata 4 V

laua`e fern (Phymatosorus scolopendria) 4 V

Hilo grass (Paspalum conjugatum) 3 V

yellow ginger (Hedychium flavescens) 4 V

taro (Colocasia esculenta) 3 0

yellow guava * (Psidium guajava) 3 F

Christmas-berry * (Schinus

terebinthifolius) 3 0

octopus tree * (Schefflera actinophylla) 3 0

Java plum * (Syzygium cumini) 3 0



Canopy- 100% shaded

Coverage Abundance

Left bank

laua`e fern (Phymatosorus scolopendria) 5 A

yellow ginger (Hedychium flavescens) 5 F

autograph tree * (Clusia rosea) 5 0

mango * (Mangifera indica) 5 0

Java plum * (Syzygium cumini) 5 0

Right bank

laua`e fern (Phymatosorus scolopendria) 3 F

yellow ginger (Hedychium flavescens) 5 A

haole-koa (Leucaena leucocephala) 5 F

autograph tree * (Clusia rosea) 5 0

yellow guava * (Psidium guajava) 5 V

octopus tree * (Schefflera actinophylla) 4 0



Vegetative coding system used in categorizing riparian vegetation.


<5% = 1 Rare (1 - 4 plants) = R

5% - 25% = 2 Occasional (5 - 14 plants) = O

26% - 50% = 3 Frequent (15 - 29 plants) = F

51% - 75% = 4 Abundant (30 - 99 plants) = A

76% - 100% = 5 Very abundant (> 100 plants) = V


An ‘o’opu-nakea (Awaous guamensis) guarding its eggs from predators.




Five fish species were found in Limahuli stream; all are native to Hawaii. Four of these were endemic (found only in Hawaii and nowhere else) and three were true gobies (with fused pelvic discs). The fourth, Eleotris sandwicensis (`o`opu-okuhe or akupa), does not have this fused pelvic disc. The significant species in this list were the indigenous goby, Awaous guamensis, (`o`opu-nakea) and the endemic goby, Sicyopterus stimpsoni (`o`opu-nopili). They were found in station II, III, and IV. There appeared to be slightly more nopili than nakea in Limahuli stream during these surveys.

The nakea supports a small commercial and recreational fishery on Kauai with Hanalei and Wainiha rivers as the prime fishing areas. The last time it was sold in the markets, two years ago, it was $10.00 a pound. This is the largest of the native gobies, reaching a minimum of over 30 cm standard length in Kauai streams. It is well known for its downstream migrations usually in association with freshets or flash floods. Spawning occurs near the mouth of rivers and streams. The best published information on this goby is Ego (1956) but there are extensive on-going studies on the nakea by University of Hawaii (Robert Kinzie), Kauai Community College (Mile Kido) and Department of Land and Natural Resources (Don Heacock) scientists.

The nopili has no commercial value at present although in the olden days it was reserved for the exclusive (food fish) use of the kahunas (priests)(Titcomb 1972), and was thought to bring good luck. According to Titcomb (1972) the largest nopili (up to 18 cm standard length) were found in Kauai streams, particularly in Wainiha, Hanalei and Makaweli. During this survey, the largest we caught was 10cm standard length. This species has been recommmended by Timbol and Maciolek (1978) as an indicator species. Its decline in population density, or in extreme cases, its disappearance in a stream is a good indication of serious habitat degradation. Extensive information on the biology of the nopili is available in Tomihama (1972) and Yuen (1986).


Native aholehole (Kuhlia sandvicensis) .

The other two fish species, both endemic, were Stenogobius hawaiiensis called `o`opu-naniha, and Kuhlia sandvicensis, known as aholehole. The naniha is a small (10 cm SL maximum) goby that lives in and around stream and river mouths. The aholehole is a marine visitor and is found in freshwater only when young; at maturity it lives in the marine environment.

Two species are listed in scientific publications as "threatened." The goby fish (`o`opu- nakea) is listed "of special concern." by Deacon, et al. (1979) and depleted by Miller (1972), but has no legal protection (Johnson 1987). The second, also a goby fish (`o`opu-nopili) is listed "of special concern" by Deacon, et al. (1979) but like the preceding species, has no legal protection (Johnson 1987). For definition of "depleted," see footnote in table; Miller's "depleted" is about equivalent to Deacon, et al.'s "of special concern.."

The gobies in this species inventory were all amphidromous, a designation for species which are migratory between fresh and salt water, having marine larval phases. The life history of this species can be generalized as follows: spawning may occur over a period of months (July-December), in the lower reaches of the streams. Hatchlings are carried to the sea by stream flow where they grow and develop over a period of between four and seven months as plankton ("as long as 160 days" according to Radtke, Kinzie, and Folsom 1988). The larvae then metamorphose into post-larvae known as hinana near the mouths of streams, settle on appropriate substrata, and migrate upstream to their places of permanent residence. This life style requires an unimpeded passageway from the stream mouth to the upper elevations. The occurrence and relative abundances of the fishes are shown in Table 7.

Table 7. The occurrence and relative abundances of the fishes in Limahuli Stream.


Scientific name Sampling Station (common name) 1 2 3 4



Awaous guamensis 0 0 ++++ ++


Eleotris sandwicensis 0 0 0 ++

(`o`opu-okuhe or akupa)

Kuhlia sandvicensis 0 0 0 ++++


Sicyopterus stimpsoni 0 ++++ ++++ ++


Stenogobius hawaiiensis 0 0 0 +



Legend: ++++ = very abundant (11 or more

+++ = abundant (6 - 10)

++ = common (2 - 5)

+ = uncommon (1)

0 = not collected or seen, possibly absent.


Table 8. Summary list of fishes occuring in Limahuli stream, Kauai (July - September 1989).

Includes common and/or local names, origin, and listing in scientific and/or official register.


Scientific Names Common Names Origin Listing



guamensis `o`opu-nakea indigenous special concern

(Deacon et al 1979)


(Miller 1972)

Eleotris `o`opu akupa or

sandwicensis `o`opu-okuhe endemic none


sandvicensis aholehole endemic none


stimpsoni `o`opu-nopili endemic special concern

(Deacon, et al.



hawaiiensis `o`opu-naniha endemic none


Terms used:

endemic = occuring naturally in Hawaii only

indigenous = occuring in Hawaii and also elsewhere

native = both endemic and indigenous

alien = brought to Hawaii either intentionally or accidentally by man

Listing: (Miller 1972 definitions)

Threatened = facing extinction, needs special protective measures

Depleted = still occurs in numbers adequate for survival but heavily depleted and continues to decline at a rate substantially greater than can be sustained.


Crustaceans (Shrimps and Prawns)

Four crustacean species were found in Limahuli stream, including three decapods, two of which are endemic to Hawaii, and one amphipod, Gammarus sp., a sideswimmer of unknown origin. One endemic, Atyoida bisulcata, (commonly referred to as mountain `opae, `opae-kala'ole or kuahiwi) is of some economic importance, as well as the alien species, Macrobrachium lar, both of which are harvested for home consumption. The other endemic, Macrobrachium grandimanus, has limited use as fish bait.


Native mountain opae kala’ole (Atyoida bisulcata).

The mountain`opae, Atyoida bisulcata, was both ubiquitous and very abundant in Limahuli Stream, except at the lowest elevation sampling station. All of the mountain `opae collected in the upper elevation (Stn I and Stn II) were adults, while those from the lower elevation (Stn III) were mostly juveniles and postlarvae; those from the lowest elevation (Stn IV) were post larvae, barely a centimeter long. At the highest elevation station (Stn I), only the mountain `opae (all adults) was present in very large quantities. For example, we obtained 379 adults in four square meters ( = 1895 individuals in 20/meter square).


Alien Tahitian prawn (Macrobrachium lar)

The alien crustacean Macrobrachium lar (tahitian prawn) is common in the lower stations III and IV, while the endemic crustacean Macrobrachium grandimanus was collected only in the lowest elevation (Stn IV).

Table 9. Crustaceans in Limahuli stream, Kauai (July - September 1989). Includes common and /or local names, origin, and listing in scientific and /or official register.


Scientific Names Common names Origin Listing




Gammarus sp. sideswimmers unknown none


1. Atyoida `opae-kala`ole or

bisulcata `opae-kuahiwi endemic none

(mountain shrimp)

2. Macrobrachium

grandimanus `opae-oeha`a endemic none

3. Macrobrachium

lar tahitian prawn alien none


Terms: endemic = occurring naturally in Hawaii only;

indigenous = occurring in Hawaii and also elsewhere;

native = both endemic and indigenous;

alien = brought to Hawaii either intentionally or accidentally by man.


Table 10. Distribution and relative abundances of crustaceans in Limahuli stream, Kauai (July - September 1989).

Scientific Name Sampling Station

(Common Name) I II III IV


Atyoida bisulcata ++++ ++++ ++++ +

(mountain `opae)



(`opae-o`eha`a) 0 0 0 +

Macrobrachium lar

(tahitian prawn) 0 0 ++++ +++


Legend: ++++ = very abundant (11 or more)

+++ = abundant (6 - 10)

++ = common (2 - 5)

+ = uncommon (1)

0 = not collected or seen, possibly absent.



Eleven species of aquatic insects were found living in or around substrate in Limahuli Stream (Table 11). These species spend part of their lives in water as nymphs, naiads or larvae. Four of these species are endemic, three are alien, and the rest are of unknown origin. Their distribution and relative abundances are in Table 12. The most important insect components were chironomids and caddisflies immatures. These form the major food source for fish and large crustaceans in the stream. Caddisfly larvae were found in all four stations sampled, but were more abundant in the lower stations. The upper stations were charcterized by an abundance of coleopteran waterpennies and caddisflies larvae.

A unique feature of the macrobenthos in Limahuli stream was the presence of water pennies (Coleoptera: Psephenidae) which were discovered inhabiting the basalt bottom of fast flowing habitat in Sampling Stations I and II (upper elevations). A single specimen was collected in Station IV in July just after a big freshet. The species was not collected in September in that same station. It may have been washed down on the earlier date by the freshet. Waterpennies have yet to be recorded in any other streams on Kauai. According to White, Brigham and Doyen (1984) the life history of these insects have not been well studied. What is known is that they lay their eggs in masses of several thousands on submerged objects in the swifter parts of riffles. They hatch into larvae which leave the water to construct pupal chambers in moist soil after which they hatch into adults. The adult stage is very short with little or no feeding.

Three insect species were found in all four stations: Cheumatopsyche pettiti (analis) (caddisfly larva), Oxyethira maya (microcaddisfly), and Megalagrion sp. (damselfly naiad). On the basis of abundance the caddisflies were more numerous in the lower stations while the waterpennies were more abundant in the upper stations.

Station II harbored the greatest number of insect species, nine out of 10. The pygmy mole cricket listed in that station was collected incidentally by electroshocking, not by surber net. The damselfly naiad was also found among the electroshocking collection but did not figure in the numbers shown in the tabulation.

Table 11. Insects in Limahuli stream, Kauai (July - September 1989). Includes common and /or local names, origin, and listing in scientific and /or official register.


Scientific Names Common names Origin Listing



Chironomidae midge larvae

1. Orthocladius

grimshawi alien (?) none

2. Calepsectra

hawaiiensis endemic none

Ephydridae brinefly larvae endemic none

Tipulidae cranefly larvae endemic none

Tricoptera caddisfly larvae

1. Cheumatopsyche

pettiti (analis) caddisfly alien none

2. Oxyethira

maya microcaddisfly alien none


Psephenidae waterpennies unknown none

Odonata: Zygoptera

Megalagrion sp. damselfly naiad endemic none

Ephemeroptera mayfly nymph unknown none


Tridactylidae pygmy mole cricket unknown none


Terms used:

endemic = occurring naturally in Hawaii only;

indigenous = occurring in Hawaii and also elsewhere;

native = both endemic and indigenous;

alien = brought to Hawaii either intentionally or accidentally by man.

Table 12. Distribution and relative abundances of aquatic macrobenthic insects in Limahuli stream, Kauai (July - September 1989).


Scientific Name Sampling Station

(Common name) I II III IV

I. Diptera (flies)

1. Orthocladius

grimshawi 0 ++++ 0 0

(midge larva)

2. Calepsectra

hawaiiensis 0 +++ 0 0

(midge larva)

Ephydridae + 0 0 0

(brinefly larva)

Tipulidae 0 +++ 0 0

(cranefly larva)

II. Tricoptera


1. Cheumatopsyche

pettiti (analis) + + +++ ++++

(caddisfly larva)

2. Oxyethira maya + + + ++



III. Coleoptera (beetles)

Psephenidae +++ ++++ 0 +


IV. Odonata

Zygoptera (Damselfly)

Megalagrion sp. + ++ + +

(damselfly naiad)

V. Ephemeroptera 0 + 0 0

(mayfly nymph)

VI. Orthoptera

Tridactylidae 0 + 0 0

(pygmy mole cricket)


Legend: ++++ = very abundant (11 or more)

+++ = abundant (6 - 10)

++ = common (2 - 5)

+ = uncommon (1)

0 = not collected or seen, possibly absent.


Limahuli Garden and Preserve.



Keep the fish population native

The fish population in the stream is entirely native and serious consideration should be given towards maintaining this now rare condition in Hawaii. The urge to stock the stream with poeciliids (i.e. mosquitofish, swordtail, mexican molly) to control mosquitoes should be resisted strongly. These alien species compete with the native `o`opus for food and living space and they are the more hardy, meaning, they will have a detrimental impact. The Hawaii Cooperative Fishery Research Unit has published information on this subject (Hathaway 1980, Norton, et al. 1978).

Visitors in the valley who are allergic to mosquito bites may be provided with mosquito repellent sprays. The historical introduction of these poeciliids to control the mosquitoes has obviously not worked.

Based on casual conversations with freshwater fishing enthusiasts, it appears that the spread of smallmouth bass (Micropterus dolomieui) and largemouth bass (Micropterus salmoides) in Kauai's streams and reservoirs is due to unauthorized stocking. These fish are good "fighters" and provide the excitement all light tackle fishermen seek. The (unauthorized) introduction of these predators will mean disaster to the endemic fish and crustacean populations in the stream. I have no suggestion as to how to control unauthorized introductions. Intentional stocking of alien fish is subject to government regulation. The Department of Land and Natural Resources, Aquatic Resources Division, can provide the specifics on this regulation.

Avoid major water diversion and channel alteration.

Due to the migratory behavior of native Hawaiian stream fauna, diversion is particularly detrimental as these species require access to the ocean as well as migratory pathways to habitat above the diversion. There are several examples in Hawaii demonstrating the adverse effects of such disruptions to migratory fauna. Stream segments in West Maui, Honokowai, Kahoma, and Waikapu do not have any native fishes and crustaceans above their diversions (Timbol and Maciolek 1978). Lau (1977) attributed lower Lentipes abundance in Piinau stream on East Maui to a partial diversion of stream flow.

Total diversion may result in the destruction of the stream ecosystem below the diversion and can render the segment above the diversion inaccessible to migratory fauna. Even partial diversion usually results in simultaneous modification of environmental conditions. For example, modification of flow rate may result in changes to the stream bed, scouring power of the stream, oxygen carrying capacity of the water temperature, and other parameters.

Avoid decreases in stream flow

Hawaiian endemic fish and crustacean species evolved in fast flowing water conditions, a characteristic of pristine Hawaiian streams. A decrease in stream flow may cause the disappearance of the endemic goby Sicyopterus stimpsoni (`o`opu nopili). An example of this was reported by Timbol (1979) in Kawa stream, a tributary of Kahana stream on windward Oahu. It is known that less oxygen (even in saturated conditions) is available to organisms when velocities are reduced. The fact that `o`opu nopili is found only in strong flowing, high-quality streams (Timbol and Maciolek 1978, Tomihama 1972) suggests that it requires high oxygen concentrations.

A decrease in flow velocity may also lead to a decrease in the insect larvae population. Insect larvae are the food resource for fish. According to Ruttner (1963) velocities below 20 cm/s will form a streambed of mineral organic mud and large quantities of organic detritus; velocities between 20 and 40 cm/s will form a substrate of small, medium, to fist-size gravel. Velocities above 120 cm/s will produce a bottom composed of large stones to boulders. The optimum velocities for invertebrate productivity, according to Ruttner (1963) are velocities between 20 and 60 cm/s. Measured velocities for Limahuli are between 3 and 55 cm/s.

Minimize the removal of riparian vegetation and habitat disturbance

Avoid the removal of streamside vegetation and if a considerable stretch of the stream bank must be cleared, it should be replanted as soon as possible. Riparian clearing may cause high insolation resulting in elevated water temperatures and excessive evaporation. Excessive evaporation could lead to reduced stream flow. Reduced stream flow means higher water temperatures. Work done by Timbol and Maciolek (1978) show that stream channels without riparian vegetative canopy have higher water temperatures than streams with intact canopy.

The effects of elevated temperatures can be divided into three categories: lethal, metabolic, and behavioral. Lethal temperatures make up the range within which the animal will die. Metabolic effects are "delayed effects" as in growth acceleration resulting in the inability to reach and/or pass a critical point in the animal's life cycle (Andrewartha and Birch 1954). Behavioral effects are the organism's responses to the environment.

Laboratory studies done by Hathaway (1979) showed that the lethal temperature for adult Awaous guamensis (`o`opu-nakea) is between (first death to final death) 37.2 and 38.8 degrees centigrade with 50% (LT50) of the fish dying at 38.1 degrees centigrade. The post-larvae (hinana) are slightly more tolerant with a range for mortality between 39.0 and 39.3 degrees centigrade with LT50 at 39.3 degrees centigrade. However, little is known about the effect of elevated temperature on the vitality of the postlarvae.

A motile animal will leave an area when conditions become unfavorable and will not voluntarily remain in the area until conditions become lethal. Thus, there will be a decrease in numbers as those that can leave will do so. Timbol and Maciolek (1978) found that altered (channelized) streams have higher water temperatures than unaltered ones. In the unaltered streams, native species were dominant in both number of species and biomass. Alien species were dominant in altered streams

Keep maintainance roads and hiking trails away from the stream

The impact of maintenance roads and hiking trails comes from the resulting erosion and siltation in the streambed. Water turbidity and excessive sedimentation will alter the character

of the stream. Burns (1972) reported turbidities greater than 3,000 ppm resulting from such construction. Excessive sedimentation may alter the biological character of the stream. Fine particulate matter will become suspended in the water increasing turbidity and decreasing light penetration resulting in reduced primary productivity. Fine particles also have the effect of clogging the gills of fish which could cause suffocation. Settling of particles in rapids and riffles will reduce the natural habitats of the economically and biologically valuable endemic residents of Limahuli stream.

Limahuli valley and stream a nature park?

It is rare that an almost pristine stream is located in a valley owned by a single entity (almost anyway). Serious effort should be expended to manage Limahuli Valley for both its botanical and aquatic animal resources. Needless to say, the endemic fish and crustaceans now found in Limahuli valley are priceless.



Andrewartha, H.G. and L.C. Birch. 1954. The Distribution and Abundance of Animals. Univ. of Chicago Press, Chicago.

Bovee. K.D. and T. Cochnauer. 1977. Development and evaluation weighted criteria probability of use curves for instream flow assessment: Fisheries. FWS/OBS-77/66. USFWS Cooperative Flow Service Group. Fort Collins, Colorado.

Burns, J.W. 1972. Some effects of logging and associated road construction on northern Colorado streams. Trans. Am. Fish. Soc. 101: 1-17.

Deacon, J.E., G. Kobetich, J.D. Williams, S. Contreras, and others. 1979. Fishes of North America: Endangered Threatened, or Of Special Concern. Trans. Am. Fish. Soc. 101: 239-252.

Elliot, M.E. and E.M. Hall. 1977. Wetlands and Wetland Vegetation. Prepared for the U.S. Army Corps of Engineers, Pacific Ocean Division. Fort Shafter, Honolulu. Contract No. DAC W -84-77-C-0014.

Fisher, S.G. and G.E. Likens. 1973. Energy flow in Bear Brook, New Hampshire: An integrative approach to stream ecosystem metabolism. Ecol. Monogr. 41: 421-439.

Hathaway, C.B. Jr. 1978. Stream channel modification in Hawaii. Part C: Tolerance of Native Stream Species to Observed Levels of Environmental Variablility. FWS/OBS- 78/18. U.S.

Fish and Wildlife Service National Stream Alteration Team, Columbia, Missouri.

Hynes, H.B.N. 1970. The Ecology of Running Waters. University of Toronto, Toronto.

Johnson, J.E. 1987. Protected fishes of the United States and Canada. American Fisheries Society, Bethseda, Maryland. 42 p.

Lau, E.Y.K. 1973. Dimorphism and speciation in the Hawaiian freshwater goby genus Lentipes. B.A.Honors thesis, Biology Department, University of Hawaii at Manoa, Honolulu, Hawaii. 83 p.

Miller, R. 1972. Threatened freshwater fishes of the United States. Trans. Am. Fish. Soc. 101: 239-252.

Minshall, G.W. 1967. Role of allochthonous detritus in the trophic structure of a woodland spring brook community. Ecology 48: 139-144.

Stream Monitoring GIS Stream Surveys Aquatic Organisms Conference Papers Islands Survey