SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September 1999
13
Indigenous ecological knowledge and its role
in fisheries research design: A case study from
Roviana Lagoon, W estern Province, Solomon Islands
by Richard Hamilton1 and Richard Walter 2
Introduction
Over the last two decades the island states of
Melanesia have experienced a rapid increase in
coastal development activities. These include the
large commercial tuna pole-and-line fisheries, ecotourism, aquaculture, coastal logging and, in
recent years, the live reef fish trade. These activities make significant contributions to local and
central economies but concerns have been raised
about their impact on local ecosystems, and their
long-term
environmental
sustainability
(Chadwick, 1999; Mathews et al., 1998; Veitayaki,
1997). In response to these concerns increasing
pressure has been put on government agencies to
develop coastal management plans and on developers to utilise these plans in the design of sustainable management practices (Olsen et al., 1997).
Unfortunately the implementation and success of
such policies has been limited. In part this stems
from the difficulties and cost of acquiring the high
quality scientific data that most management
plans and monitoring programmes rely on. As
Johannes et al. (1993:1) point out for coastal fisheries, ‘The cost and complexities of effectively
monitoring and managing small, multi-species,
multi-method reef and lagoon fisheries along conventional lines have generally proven prohibitive’.
Nevertheless, in many cases extensive ecological
information on the relevant coastal environments
already exists within the knowledge base of the
local communities. Further, these same communities are often operating management and monitoring programmes of sorts, within the context of
customary marine tenure (CMT) systems (Foale,
1998a; Foster & Poggie, 1993; Hviding, 1991;
Johannes, 1981).
This cultural information, often referred to as traditional ecological knowledge (TEK), is widely
recognised as potentially useful to fisheries
researchers and managers, especially when used in
conjunction with conventional scientific data,
‘Local knowledge combined with specialised
knowledge of the outside researcher is considered
by advocates of the PAR [participatory actionresearch] to be more potent than either knowledge
1. Department of Marine Science, University of Otago, New Zealand
2. Department of Anthropology, University of Otago, New Zealand
alone in understanding reality’ (Christie & White,
1997:172). However, indigenous cultural information is useful in coastal management for more than
its role in providing the baseline ecological data
required for planning purposes. Frequently, development projects are set up within the territories of
indigenous communities and those communities
are directly and indirectly involved on a number of
different levels (Alcala, 1998; Ruddle et al., 1992). In
these cases the successful operation of development projects depends on a proper understanding
of local political and subsistence economies, ideology and social structures. In fact, the failure of
development projects in the Pacific is probably
more commonly the result of a failure to understand local cultural systems than of inadequacies
in ecological, technological or market research.
Unfortunately, indigenous knowledge is usually
either ignored or used inadequately by fisheries
researchers. There are two fundamental problems.
The first is that TEK and other types of indigenous
knowledge exist as inseparable parts of complex
cultural systems, and it requires anthropological
methods to describe and interpret this information
in a meaningful manner. These skills, however,
are usually difficult, time consuming and well
beyond the professional training of most fisheries
scientists, resource planners and project managers
working in island Melanesia (Christie & White,
1997; Clark & Murdoch, 1997). The second closely
related problem, is that frequently, when scientists
have attempted to incorporate indigenous knowledge into their baseline research, the result has
been a naive reporting of interview data or observations taken out of their cultural and historical
context. The results are often misleading and misrepresentative. ‘The romantic and uncritical
espousal of traditional environmental knowledge
and management is an extreme almost as unfortunate as that of dismissing it’ (Ruddle et al.,
1992:263). A particularly good example of this is
the common practice of interpreting CMT systems
as conservation strategies. Although CMT systems
can have conservation outcomes, it has recently
been demonstrated that in most cases this is a side
effect of systems designed primarily for ‘gain, not
14
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September1999
restrain’ (Ruttan, 1998; Aswani, 1998; Polunin,
1984). An integrated approach to coastal resource
planning is required, in which the specific skills of
anthropologists and fisheries scientists are properly utilised. In particular, cultural information
needs to be gathered systematically and treated
with the same critical scrutiny that is applied by
scientists to any other data set.
In this article we take three premises as our starting point:
1. TEK contains complex ecological information
obtained through generations of observation
and experience;
2. TEK is culturally structured and if it is to be
taken out of its original context and used in an
empirical western scientific framework, its
appropriateness to this framework should be
established through systematic testing; and
3. One of the most powerful ways in which TEK
can be used in fisheries or other types of ecological research is in the establishment of
research designs.
We present below a case study in the development
of a model for testing TEK, and in establishing the
parameters within which it can be used in a scientific research design. The work is situated in
Roviana Lagoon, on the island of New Georgia,
Western Province, Solomon Islands. Roviana
marine ecologies and local communities have
Pap
u
already felt the impact of development activity in
the form of forestry, and of the Solomon Taiyo
tuna and baitfish industry (Nichols & Rawlinson,
1990). Several other small-scale coastal ventures
have been set up over the last few years, and more
are planned for the near future. In this study, we
examine local knowledge of aggregating
behaviours of the Carangidae fishes in a confined
part of the lagoon ecosystem. Carangids, or trevally, are a diverse family of reef and lagoon fish
which are important in the local subsistence and
artisanal fishery. In this study, detailed ethnographic data on Carangidae behavior were collected from local informants using traditional anthropological techniques of interview and participant
observation. This information served as a basis for
designing hypotheses on the tidal movements and
aggregating behaviours of Carangidae within
Roviana Lagoon, which were then tested using scientific field observations. The purpose was not to
test the accuracy of local ecological knowledge,
but to show how it can assist in the development
of fisheries research designs.
Environmental background
The Solomon Islands lie between 7° and 10°S and
extend 1400 km eastward from New Guinea
(Figure 1). The islands display remarkable diversity in both terrestrial and marine environments.
The dominant island forms are volcanic, but
examples of atolls and raised reef islands are
also present. Extensive lagoon systems occur in
Western Province.
.... .. Bismarcks
New Britain
a-N
ew
G
uine
..
a
...
Solomon
Is
.
..
New
Hebrides
Samoa
Fiji
Cook Is
..
Society Is
Tonga
New Caledonia
Austral Is
Australia
40°
40°
New
0
Zealand
150°
170°
1000km
150°
Figure 1. The Solomon Islands in the Pacific (Survey and Mapping Division, Honiara)
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September 1999
The New Georgia Group, in Western Province, is
made up of nine main islands that extend for
approximately 270 km. The largest island is New
Georgia which is fringed by Roviana Lagoon to
the south-west and Marovo Lagoon to the northeast (Figure 2).
Roviana Lagoon is a body of shallow water
approximately 50 km long enclosed between the
New Georgia mainland and a series of uplifted
coral reef islands lying 2 to 3 km offshore. The
lagoon supports a wide range of micro-environments including: estuaries, mangrove forest, river
mouths, mudflats, coral atolls, barrier reefs,
marine lakes and sea-grass beds (Aswani,
1997:245). Each of these, as well as the adjacent
offshore zones, provides unique opportunities for
the fishing communities of Roviana Lagoon. A
number of major passages connect the open sea to
the lagoon. These are remnants of earlier river systems that flowed through the now-flooded coastal
plains. These passages contain a relatively high
biodiversity and are consequently of especial significance in local fishing systems. Their rich biodiversity occurs partly because they connect several
different ecological zones and many species utilise
them to move between feeding areas.
Roviana Lagoon is occupied by a number of
closely related tribal groups occupying a dozen
or so large villages scattered along the mainland
and barrier islands as well as a number of smaller residential clusters. Roviana is the most
homogeneous linguistic region of New Georgia,
where the Austronesian language of Roviana is
spoken in all the villages. Elsewhere in New
Georgia the 30,000 inhabitants speak at least
seven major languages, including several nonAustronesian languages. The Roviana subsistence economy is based on shifting horticulture
with an emphasis on the recently introduced
sweet potato and cassava, as well as the traditional crops: taro, yam, banana and sago. The
main protein intake is from marine sources
(Aswani 1997:189) but pig and chicken husbandry and small-scale hunting also takes place.
Today, these subsistence activities are often subsidised by waged labour and copra production.
Carangids were selected for this case study for
several reasons. First, the ecology of carangids in
tropical settings is poorly studied (Sudekum et al.,
1991). Second, the carangids are important in the
Roviana subsistence fishery and thus the local
body of knowledge was expected to be detailed
and accessible. Third, there are potentially important trophic interactions between the Taiyo baitfishery and carangids which were investigated as
part of the field research, but are not reported here
(see Hamilton, 1999).
The Carangidae is a family of strong swimming,
open water carnivorous fishes known variously as
trevally, kingfish, queenfish, jack, scad, mackerel,
and rainbow runners (Gunn, 1990; Randall, 1990).
Their habitats range from brackish water estuaries
to deep offshore reefs, with a few species being
pelagic in the open ocean (Holland et al., 1996). At
least 17 species of carangid are present in Roviana
Lagoon, and all are important food fishes. In this
VELLA
LAVELLA
NEW GEORGIA GROUP
BAGHA
KOLOBANGARA
NEW
GEORGIA
RANOGA
15
PARARA
RENDOVA
VANGUNU
BULO
GATOKAE
Figure 2. The New Georgia Group (Survey and Mapping Division, Honiara)
16
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September1999
study we concentrate on three species, Bluefin
trevally (Caranx melampygus), Golden trevally
(Gnathanodon speciosus) and Bigeye trevally (Caranx
sexfasciatus). with additional observations on other
members of the genera Caranx and Carangoides.
Study area
The research was carried out in Honiavasa passage from a base in Nusabanga village, in the
western part of Roviana Lagoon (Fig 3).
Nusabanga village is a sister village to Sasavele
village, located on the western bank of Honiavasa
Passage. The two villages share very close ties,
often participating in common fishing trips and
other economic activities.
The field work reported here is a part of a larger
project on fish ecology and indigenous subsistence
practices carried out in 1996–1997. The full results
of the complete project will be reported elsewhere.
The field component of this study had two parts.
The first involved gathering indigenous information on the behavioural ecology of Carangidae in
Roviana Lagoon. This information forms the basis
on which the traditional Roviana Carangidae fishery is structured. Using this data, a set of hypotheses linking environmental variables with
Carangidae behaviour was established. The second part of the research involved testing these
hypotheses using standard field survey techniques.
Ethnographic methods and results
Honiavasa Passage is one of the main deep water
passages connecting the inner lagoon to the outer
sea. It is approximately 600 m across at the sea
entrance and narrows to around 150 m at its narrowest point. The coral walls of the passage are
nearly vertical, and descend to a maximum depth
of 42 m. The passage environment is characterised
by mainly dead corals to a depth of 15 m, although
the submarine coral walls support large and
diverse migrant vertebrate communities.
The field component of the research discussed
here was carried out from early- August until lateOctober, 1997. During this period one of the
authors (RH) resided at Nusabanga village, where
he participated in the daily life of the village and
worked regularly with the local fishers, to gain as
wide an understanding as possible of the local
fishing system. This followed the viewpoint that
empirical facts about biology or ecology can not be
Figure 3. The western part of Roviana Lagoon, showing Honiavasa Passage, Munda township
and Nusabanga (Mbanga) and Sasavele village (Survey and Mapping Division, Honiara)
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September 1999
extracted from TEK without an understanding of
the structure and ideology of the wider cultural
and subsistence systems of which it forms a part.
During the fieldwork period records were taken of
51 fishing trips. Information was recorded on
Catch Per Unit Effort (CPUE) and on the technological, meteorological and environmental details
of each major fishing trip. Basic information on
indigenous taxonomy and fish names were
acquired through formal interviews conducted in
Solomon Island Pidgin, the local lingua franca.
Local fish names were collected and then cross
checked against standard fish guides (Allen &
Swainston, 1992; Randall et al., 1990). Local informants were also asked to identify species seen on
underwater video footage taken in Honiavasa
Passage. In addition, formal interviews were conducted with 11 fishers from Nusabanga and three
from Sasavele. These persons were selected
according to their recognised status as fishing
experts within their respective villages. All interviews were conducted in both Roviana and
Solomon Island Pidgin, and were recorded on
videotape. Each fisher was asked to identify his
main fishing locations, fishing methods used and
the target species. Particular attention was paid to
the spatial, temporal and technological structure
of Carangidae fishing practices.
Hviding (1996), and also to that Nggela
(Foale,1998b). All these systems differ considerably from Linnaean taxonomy, in that the level of
classification reflects the value of the fish within
the subsistence system, as well as basic morphology. For example, the small Damselfishes
(Pomacentridae) are not eaten in Roviana and are
all referred to using the generic term kipa.
Although Roviana people are well aware of the
variation within the Pomacentridae family, no
further attempt is made at finer taxonomic discrimination. One the other hand, some members
of the family Scaridae are very important in the
subsistence economy, and this is reflected in their
taxonomic treatment where different names are
assigned to different size classes (Aswani,
1997:425). The actual names assigned to a fish
taxa often reflect behavioural characteristics
shown by that fish.
At least 17 species of Carangidae are recognised as
taxonomically separate within the Roviana system. These names are listed in Table 1 below. Mara
is a generic name covering a range of Carangidae
species, each of which is additionally named using
a more specific suffix. For example, the Orange
spotted trevally (Carangoides bajad) is named mara
liu. Liu (derived from Solomon Island Pidgin)
translates as ‘loiter’, thus C. bajad is named ‘the
trevally that loiters’, owing to its commonness in
Roviana Lagoon.
On the basis of the field participation results and
interview data we can define the basic parameters
of indigenous fish taxonomy
and
the
Table 1. Local Carangidae taxonomy
Roviana fishing calendar. For the most part,
we observe only minor
Indigenousname Commonname
deviation from the
more detailed study
recently reported in
Mara
Generic: Trevally
Aswani (1997). We can
Balubalu
Rainbow runner
also describe aspects
Ganusu
Smooth tailed trevally
of Roviana folk ecoloLasilasi
Double-spotted queenfish
gy of Carangidae, with
(and perhaps other
queenfish species)
emphasis on tide related movements and
Laqu belama
Tille trevally
lunar aggregations in
Mara batu batu
Griant trevally
Honiavasa Passage.
Mara batu papaka
These results are sumMaaru hipu gele
Gold-spotted trevally
marised below and
Mara hobu
Blue trevally
reported in more
Mara labe
Pennantfish
detail in Hamilton
Mara lamana
Bluefin trevally
(1999).
Taxonomy
The structure of
Roviana fish taxonomy is almost identical
to that of Marovo,
well described by
Mara liu
Mara madali
Mara popana
Mara roba
Moturu
Moturu kove
Paki pakete
17
Orange-spotted trevally
Bump-nosed trevally
Golden trevally
Snub-nosed dart
Bigeye trevally
Black trevally
Black-spotted dart
Scientific name
Carangidae
Elagatis bipinnulata
Selaroides leptolepis
Scomberoides lysan
Caranx tille
Caranx ignobilis
undetermined
Carangoides fulvoguttatus
Carangoides ferdau
Alectes ciliaris
Caranx melampygus
Carangoides bajad
Carangoides hedlandensiis
Gnathanodon specious
Trachinotus blochii
Caranx sexfasciatus
Caranx lugubris
Trachinotus bailloni
18
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September1999
The Roviana lunar cycle
The lunar cycle plays a major role in the Roviana
fishing system since there is a well-understood
correlation between lunar phase and Carangidae
(and other species) behaviours which can be
exploited by local fishers. Many of the days of the
lunar month are given specific names. A version
of the Roviana lunar calendar collected in
Nusabanga is included as Table 2. This example is
based on a 30-day lunar month. A more detailed
version has been collected by Aswani (1997:238)
and Hviding (1996) provides a similar calendar
from Marovo Lagoon.
Tidal movements of Carangidae
According to local informants, Carangidae movements within Roviana Lagoon correspond to tidal
flows. The following species of Carangidae are
said to move into the inner lagoon regions
through Honiavasa Passage on the rising tide, and
then move back into the passage and outer reef
areas on the falling tide: Carangoides ferdau,
Carangoides fulvoguttatus, Carangoides bajad, Caranx
tille, Caranx lugubris, Caranx sexfasciatus, Caranx
melampygus and Gnathanodon speciosus. Informants
further state that these predictable movements are
due to a predator-prey interaction. The carangids
are following their prey species, the very small
baitfish named hinambu (members of the families
Engraulidae, Clupeidae, Apogonidae, and
Atherinidae) and the large baitfish, kutukutu
(Herklotsichthys quadrimaculatus).
The baitfish are said to move up into the inner
lagoon to spawn among the mangrove roots on
the high tide, and retreat back to sea on the falling
tide, when the shallow inner lagoon areas become
exposed. Thus the tides are considered to be one
of the most important factors regulating the likelihood of a successful catch:
Before I go fishing I must watch the
movement of the sea, there are two good
times for fishing, pado gore when the tide
starts to fall, and pado sage when the tide
starts to rise, close to flood or ebb tide the
fish do not eat (pers. comm., Harry
Kama, Nusabanga village, August 1997.
Translated from Solomon Island Pidgin
by the authors).
These two tidal terms, pado gore and pado sage,
refer to the first 3–4 hours after flood and ebb tide,
respectively. Nine of the fourteen fishers interviewed stated that the best times to catch
carangids are during pado gore and pado sage.
The state of the tides not only influences the timing of a fishing event, but also its precise location
within the lagoon:
At high tide the baitfish move up into the
mangrove areas, and they are followed
by the trevally. So at high tide I fish for
trevally in the inner lagoon, near the sea
grass in the deep areas close to the mangroves. At low tide the baitfish move out
of the mangrove and go to the passage,
so I go and fish for trevally there (pers.
comm. Diliva Dava, Nusabanga village,
August 1997. Translated from Roviana by
Gaudry Kama).
Eight of the fourteen fishers interviewed also
made the observation that on the rising tide
carangids often swim near the surface in the passage, but on the falling tide they dive deeper into
Table 2. The Roviana lunar cycle
Roviana term
Lunar phase
Translation
Taloa Sidara
Tada Keke
Tada Karua
Tada Ngeta
Tada Made
Tada Lima
Tada Onomo
Tada Zuapa
Noma Sidara
Hobe Rimata
New moon
Day 1
Day2
Day 3
Day 4
Day 5
Day 6
Day 7
Days 8-14
Full moon
Taloa : No Sidara : Moon
Tada : refers to the first quarter, when the moon is first
seen in the West and drops quickly below the horizon
(before midnight).
Pae
Days 16-30
Noma : Big Sidara : Moon
Hobe : Change Rimata : Sun. The moon changes the sun.
As soon as the sun sets it is replaced by the moon.
Pae : "it is dark before the moon rises". The moon rises after 12 pm
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September 1999
the channel. Because of this, carangids are harder
to catch by trolling or casting in the passage on the
low tide, and alternative techniques must either be
adopted, or other fishing locations selected.
Lunar-related aggregations
The subsistence fishers of Nusabanga and
Sasavele recognise two main periods of the lunar
cycle as ideal times for catching carangids. The
first of these ‘Carangidae seasons’ is during the
early part of the first quarter of the lunar cycle,
and the second is around the full moon:
Mara have seasons, so whenever we want
to catch mara we must wait for its time. If
you fish for mara when it is not the right
time, you will not have much success
(pers comm. Harry Kama, Nusabanga
village, August 1997. Translated from
Solomon Island Pidgin by the authors).
Thirteen of the fourteen fishers interviewed gave
very specific and consistent information relating
to passage and inner lagoon fishing in the first
quarter of the lunar cycle. On Tada keke, Tada karua
and Tada ngeta, fishing for Carangidae is highly
productive, with all species being readily caught.
Twelve of the fourteen fishers interviewed also
recognised Tada zuapa as being a good time for
catching mara batubatu (Caranx ignobilis). C. ignobilis appear within Honiavasa Passage of Roviana
Lagoon several times during the first quarter of a
lunar cycle between the months of June through to
December.
You can catch mara batubatu on Tada keke,
Tada karua and Tada ngeta. From Tada
made till Tada onomo you will not catch
batubatu. On Tada Zuapa you will catch
batubatu, and they will come in groups
(pers comm., Simon Bae, Sasavele village,
August 1997. Translated from Roviana by
Gaudry Kama).
In the early hours of the morning on Tada zuapa, C.
ignobilis are said to aggregate in large schools
along the walls of Honiavasa Passage. The largest
of these aggregations is reported to occur off the
point of Honiavasa Island, where large groups
gather in the shallow reef area on the eastern side
of the channel, close to Site 3 (see below).
In the early morning of Tada zuapa, before
the sun is up, they start to arrive in
groups. They don’t all come in at once,
you will see one group arrive then 5-10
minutes later another group will join
them, then another group, and so on. I
have never tried to count them, but in the
19
end there is often over 100 fish. They
aggregate between Honiavasa Island
and Nitara Island, on the right-hand side
of the passage. They also move up the
left-hand side of the passage, but they
do not go up as far as Vonovonoto
Island, instead they will move into the
small passage that breaks off before
Vonovonoto Island. By the time the sun
is up the schools will have gone (pers
comm., Simon Bae, Sasavele village,
August 1997. Translated from Roviana
by Gaudry Kama).
All the persons who recognise these aggregations
believe that they form as a result of a feeding
opportunity when schools of medomedo (Siganus
argenteus) and suliri (Nematalosa come) form. Jack
Kari and Harry Kama both said that during Tada
zuapa groups of C. ignobilis travel far up into the
inner lagoon to feed, often travelling right up into
the rivers which empty into the lagoon. This
observation is consistent with reports from the
Philippines, where C. ignobilis have been caught
far up rivers (Manacop, 1952, cited in
Westernhagen, 1974).
The second ‘Carangidae season’ occurs around
full moon, a time which all the fishers interviewed recognised as being particularly good for
carangid fishing. During this lunar stage C. sexfasciatus and C. lugubris are said to aggregate in
the lagoon, often in schools. These species are
predominantly targeted at night by droplining
using ganusu (Selaroides leptolepis) for bait, or by
casting non-weighted lines with hinambu (small
baitfish) as bait. C. fulvoguttatus, the gold-spotted
trevally, is also said to aggregate in schools during this lunar stage.
In addition to these two main ‘Carangidae seasons’ other minor periods of aggregation are
known to local fishers, and exploited accordingly.
Six of the fishers stated that mara batu papaka
(unidentified Carangidae species) aggregate within the lagoon in large schools after the full moon,
during Pae. Simon Bae of Sasavele stated that large
schools arrive mainly during the last quarter, and
that during this time they travel far up into the
inner lagoon. It is also important to note that some
of the Carangidae species that inhabit the lagoon
are reported by local informants not to have a specific lunar season. For example, C. bajad is said to
be constantly present within the lagoon, regardless of the lunar stage.
Marine survey methods and results
On the basis of the ethnographic survey we can
define three testable hypotheses relating to
20
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September1999
Carangidae behaviour in and around Honiavasa
Passage:
1. Carangidae species move in and out of the passage with the tides;
2. C. sexfasciatus will be at their highest densities
within the passage around full moon; and
3. C. ignobilis will aggregate in schools alongside
Honiavasa Island before sunrise on Tada zuapa
To test these hypotheses a dive survey was carried out using Instantaneous Area Point Counts
(IAPC) during flood and ebb tides at six sites in
Honiavasa Passage. For the purpose of site selection, the passage was divided into three areas:
Outer, Middle and Inner. Two sample sites were
allocated within each area, one on the left-hand
and one on the right-hand side of the passage
(Fig 4). The three separate areas were chosen in
order to determine if distance from the open sea
influenced the distribution and abundance of
fish. Two sites were allocated within each area in
order to determine whether passage side had a
significant effect on the abundance or movements
of Carangidae.
This survey spanned two lunar cycles and was
conducted using standard IAPC visual census
methods. IAPCs were recommended
by Thresher and Gunn (1986), who
assessed the usefulness of many underwater visual transect and point count
techniques for estimating biomass and
abundance of Carangidae. The IAPC
method used here involved underwater
visual census carried out by two divers
at selected times, locations and depths
using SCUBA. In addition, underwater
video footage shot from a fixed station
was taken for identification and reference purposes. A total of 48 hours of
video was recorded. A fibreglass canoe
powered by a 15 hp motor was used
during the field surveys.
On arriving at a site, visibility was
determined in order to define differences in average visibility between the
six study sites. At a depth of 5 m a plastic silhouette of a fish (49 cm in length)
was located on the bottom. One end of
a fibreglass tape measure was attached
to the silhouette, a diver then swam
horizontally away from the silhouette
reading the distance off the tape measure at which the silhouette became difficult to distinguish. Both divers then
stationed themselves with their backs
to the reef wall and fifteen point counts
were carried out over 15 minutes. At
60-second intervals, the entire 180-degree area of
the passage visible to the diver was scanned (a
procedure taking 5-10 seconds) and all individuals in that area were recorded by species.
Observations of the number of individuals of a
given species, their length and their direction of
movement (up or down the channel) were
recorded. The amount of error in the length estimates was calculated throughout the field observations by periodically recording the ‘estimated’
and ‘actual’ length of coral heads. A total of 120
dives were carried out during the months of
September and October 1997. These consisted of
twenty dives at each site, ten on the falling tide
and ten on the rising tide.
The analysis of the IPAC data looked for interactions between the following five variables:
1. State of tide (rising, falling);
2. Direction of fish movement (in, out);
3. Lunar stage (new moon, first quarter, full
moon, second quarter);
4. Area (Outer, Middle, Inner); and
5. Side of passage (left, right).
Presence-absence data for the three most frequently sighted species of Carangidae, C. melampygus,
G. speciosus and C. sexfasciatus, were analysed
5
6
4
3
2
1
0
Figure 4. Aerial photo of Honiavasa Passage
showing the location of the six sites
1km
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September 1999
using a logistic linear model in the software package JMP™ on a Macintosh™ computer. A model
of best fit was determined for each data set, then
the significant terms in the chosen model were
examined in detail.
Total abundance data for C. melampygus were
transformed from dive log records using the standard log 10 transformation. These abundance data
were then analysed using a general linear model,
ANOVA, involving all main effects and interactions. This was carried out using DataDesk™ software on a Macintosh™ computer.
Carangidae movements with respect
to tidal flows
Log abundance (x+1) of total seen
The two most commonly sighted species of
carangid, C. melampygus and G. speciosus, both
showed a strong tendency to move in the direction
of tidal flow. Analysis of the C. melampygus log
data revealed a significant two way interaction
between tidal flow and direction of fish movement
(DF = 1, SS = 1.03, MS = 1.03, F-ratio = 11.76 and P
value < 0.01) (Fig. 5).
Analysis of G. speciosus presence/absence data
also revealed that the movements of fish within
Honiavasa Passage correlate strongly with tidal
flows (LR Statistic = 50.56, 1 DF; P < 0.01) (Fig. 6).
Lunar aggregations of Carangidae
C. sexfasciatus was the third most commonly seen
Carangid in the dive survey. Presence/absence
data analysis revealed that lunar stage influences
their relative abundance within the passage (LR
Statistic = 8.33, 3 DF; P = 0.04). C. sexfasciatus were
20
18
16
14
12
Moving in
10
Moving out
8
6
4
2
0
Rising
Falling
Tidal state
Figure 5. Log value of the total number of Caranx melampygus seen moving in and
out of Honiavasa Passage on 60 rising and 60 falling tides (Error bars = ±1 SE)
60
Percentage of time seen
50
40
30
Moving in
Moving out
20
10
0
Rising
21
Falling
Tidal state
Figure 6. The percentage of time that Gnathanodon speciosus were seen moving in
and out of Honiavasa passage on 60 rising and 60 falling tides (Error bars = ±1 SE)
22
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September1999
potential relevance in coastal resource management and development planning. Through anthropological field techniques we were able to acquire
a ready-made model of these behaviours which
could then be quickly and easily tested using standard marine science field methods. Through participant observation and interviews we identified
two specific categories of Carangidae behaviour
well understood by Roviana fishers. These categories pertained to the temporal and spatial patterning of fish activity at different scales. First,
there was the relationship between tidal change
and Carangidae movement. Second, the relationship between fish aggregating behaviour and the
lunar cycle. We were then able to create a research
design that would maximise the likelihood of us
observing these predicted behaviours. In this case,
the dive survey results verified local statements on
marine ecology. Other, more complex, research
questions are likely to reveal a more complex relationship between traditional and scientific knowledge systems.
seen the highest percentage of the time (15%) during the full moon lunar phase (Fig. 7). In the new
moon and first quarter they were seen approximately 5% of the time and during the second quarter they were seen only 2% of the time.
There were insufficient sightings of C. ignobilis and
mara batu papaka to warrant statistical analysis, but
observations made during the dive survey and
during other field exercises support local predictions about their aggregating behaviours. C. ignobilis were seen only once out of 120 survey dives,
and this sighting occurred on Tada zuapa the lunar
day that C ignobilis schools are reported by local
fishers to form within Honiavasa Passage. The
sighting occurred at Site 6 in the late afternoon of
10 Sep. 1997 on a rising tide. Further, the predicted
aggregations in the early morning of Tada zuapa
were observed the following month, on the 10th of
October. The CPUE data also demonstrated C
ignobilis aggregations around Tada zuapa.
Mara batu papaka were sighted only on three occasions out of a total of 120 survey dives. These
sightings all occurred during Pae, the period following full moon. Again, this accords well with
CPUE data and with the statements made by
informants during the ethnographic survey (see
above).
In summary, we review in brief a number of benefits and motives for researchers to consider indigenous knowledge systems, and conclude by reiterating the argument for an holistic approach to
coastal research in Melanesia.
1. TEK contains baseline information on local
ecologies including information on what is present in the local ecosystems, and its temporal
and spatial patterning.
2. By drawing on TEK at the early stage of
research, scientists are able to develop testable
models, and so target their research and efficiently budget their time and other resources.
Discussion
As this study demonstrates, TEK systems contain
knowledge that is relevant to many of the types of
question asked by fisheries research scientists. In
this case we were interested in carangid movements and aggregation patterns, a question of
Percentage of time seen
30
25
20
15
10
5
0
New moon
(n = 34)
1st quarter
(n = 42)
Full moon
(n = 17)
2nd quarter
(n = 27)
Lunar stage
Figure 7. The percentage of time that Caranx sexfasciatus were seen in Honiavasa
Passage during the four lunar stages (Error bars = ± 1 SE)
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September 1999
3. TEK and CMT evolved over generations of
direct interaction between human communities
and their environment and are likely to contain
information and ideas not currently contained
within scientific models (Lalonde & Akhtar,
1994).
4. If properly understood, there is a potential for
resource managers to use CMT systems directly to manage stocks in development projects
(Hviding, 1996; Ruddle et al., 1992).
Conclusions
All knowledge systems, including scientific knowledge systems, are culturally structured and there
are potential dangers in extracting information
from one system, and applying it in another. For
that reason, research scientists working with TEK
should attempt to develop a holistic understanding
of those systems, so as to understand the context of
the information they hold. In coastal research, this
would include information on folk taxonomy and
ecology, fishing technologies and the social and
symbolic context of fishing activities.
Knowledge is often stratified by gender, age and
geographical location (Hviding, 1996; Christie &
White, 1997). For example, in Roviana Lagoon,
women fishers hold traditional knowledge on
mangrove clams, as it is almost exclusively
women who collect these food species. It would
thus be difficult for male researchers to gain
access to that information, and potentially misleading if they were to ask only male members of
the community about the exploitation of mangrove clams (although they would undoubtedly
get an answer).
We also note that in Roviana, knowledge pertaining to specific families of fish is sometimes restricted to fishers who specialise in targeting those
species. Several of the fishers in this study when
asked questions on the ecology of Carangidae
made statements similar to that of Enele Garata:
‘Oh no, I don’t know about mara, I only know
about pazara (Serranidae) and pipo (Sphyraenidae),
ask the old man Simon Bae at Sasavele, he knows
about mara’ (pers comm., Enele Garata,
Nusabanga village, August 1997. Translated from
Solomon Island Pidgin by the authors).
It is also important to appreciate that most
indigenous knowledge of marine ecologies is
ultimately directed towards identifying patterns
which maximise capture success. Thus some
details of fish behaviour which may be irrelevant
to a local knowledge base (since they have no
influence on subsistence practice) may be of
immense importance to a marine biologist studying reef ecology. Further, whereas indigenous
23
knowledge of fish behaviour will often be very
accurate, local explanations for the mechanisms
underlying these behaviours may not be compatible with scientific paradigms.
In Nusabanga, local fishers believe that mara movements through the passage relate to predator–prey
relationships with baitfish (see above). Our
research shows, however, that while the predicted
movements of fish is accurate, the explanation
probably varies amongst species (Hamilton 1999).
Ruddle et al., (1992:262) make a similar point, ‘In
some places declining yields may be attributed to
sorcery or a failure to propitiate the gods’.
In conclusion then, indigenous knowledge is an
enormously valuable resource base but one which
is often treated in a far too casual manner by
research scientists. We advocate an holistic
approach incorporating the strengths of anthropological and marine science methods and techniques. In a cross-disciplinary approach, each discipline provides the other with a degree of
accountability and more accurately represents the
knowledge base of the indigenous peoples of
Melanesia who have been living and acting in
these landscapes for up to 30,000 years.
Acknowledgements
We would like to acknowledge a number of people for their support during various phases of the
project. In the field, the assistance of Bailey Kama,
Gaudry Kama and other members of the
Nusabanga community is gratefully acknowledged. We would especially like to thank the
paramount chiefs, Chief Joseph Kama and Chief
John Roni, for giving us permission to work in the
area. Mariana and Dave Cooke (Solomon Sea
Divers) provided equipment and Johnson Seeto
provided expert guidance with taxonomy. Finally,
we thank the New Georgia Archaeological Survey
who provided logistical field support.
References
ALCALA, A.C. (1998). Community-based coastal
resource management in the Philippines: a
case study. Ocean and Coastal Management
38: 179–186.
ALLEN, G.R. & R. SWAINSTON. (1992). Reef Fishes of
New Guinea. Kristen Press INC, Madang:
Christensen Research Institute.
ASWANI, S. (1997). Customary sea tenure and artisanal fishing in the Roviana and Vonavona
Lagoons, Solomon Islands: The evolutionary
ecology of marine resource utilization. Ph.D.
dissertation, University of Hawaii, Hawaii.
24
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September1999
A SWANI , S. (1998). Patterns of Marine Harvest
Effort in Southwestern New Georgia,
Solomon Islands: Resource Management or
Optimal Foraging? Ocean & Coastal
Management 40, 2–3: 207–235.
BLABER, S.J.M., D.A. MILTON & N.J.F. RAWLINSON.
(1990). Diets of the Fishes of the Solomon
Islands: Predators of tuna baitfish and trophic
effects of baitfishing on the subsistence fishery. Fisheries Research 8: 263–286.
C HADWICK , D. (1999). Coral in peril. National
Geographic Magazine 195(1): 30–37.
CHRISTIE, P. & A.T. WHITE. (1997). Trends in development in coastal area management in tropical countries: from central to community orientation. Coastal Management 25: 155–181.
CLARK, J. & J. MURDOCH. (1997). Local knowledge
and the precarious extension of scientific networks: a reflection on three case studies.
Sociologia Ruralis 37(1): 39–60.
FOALE, S. (1998a). Assessment and management of
the trochus fishery at West Nggela, Solomon
Islands: an interdisciplinary approach. Ocean
& Coastal Management 40: 187–205.
FOALE, S. (1998B). What’s in a name? An analysis of
the West Nggela (Solomon Islands) fish taxonomy. Traditional Marine Resource
Management and Knowledge Information
Bulletin 9: 3–19.
FOSTER, K. & J. POGGIE. (1993). Customary marine
tenure and mariculture management in outlying communities of Pohnpei State, Federated
States of Micronesia. Ocean and Coastal
Management 20: 1–21.
GUNN, J.S. (1990). A Revision of Selected Genera of
the Family Carangidae (Pisces) from
Australian Waters. Records of the Australian
Museum: Supplement 12.
HAMILTON, R. (1999). Tidal movements and lunar
aggregating behaviours of Carangidae in
Roviana Lagoon, Western Province, Solomon
Islands. MSc dissertation, University of
Otago, Dunedin.
HOLLAND, K.N., LOWE, C.G. & B.M. WETHERBEE.
(1996). Movements and dispersal patterns of
blue trevally (Caranx melampygus) in a fisheries conservation zone. Fisheries Research 25:
279–292.
HVIDING, E. (1991). Traditional Institutions and
their role in the contemporary coastal
resource management in the Pacific Islands.
NAGA Vol. 14: 3–6.
HVIDING, E. (1996). Guardians of Marovo Lagoon.
practice, place, and politics in maritime
Melanesia. Honolulu. University of Hawaii
Press.
J OHANNES , R.E. (1981). Words of the lagoon.
Fishing and marine lore in the Palau district
of Micronesia. Berkeley: University of
California Press.
JOHANNES, R.E., K. RUDDLE. & E. HVIDING. (1993).
The value today of traditional management
and knowledge of coastal marine resources in
Oceania. Workshop: People, Society, and
Pacific Islands Fisheries Development and
Management ( Noumea, New Caledonia). 1–7.
L ALONDE , A. & S. A KHTAR . (1994). Traditional
knowledge research for sustainable development. Nature and Resources 30(2): 22–28.
MATTHEWS, E., J. VEITAYAKI & V.R. BIDESI. (1998).
Fijian villagers adapt to changes in local fisheries. Ocean and Coastal Management 38:
207–224.
N ICHOLS , P.V. & N.J.F. R AWLINSON . (1990).
Development of the pole-and-line fishery in
the Solomon Islands with reference to the
baitfishery and its management. Tuna
Baitfishing In The Indo-Pacific Region,
ACIAR Proceedings No. 30: 30–44.
Oslen, S., J. Tobey & M. Kerr. (1997). A common
framework for learning from ICM experience.
Ocean and Coastal Management 37(2):
155–174.
POLUNIN, N. (1984). Do traditional ‘reserves’ conserve? A view of Indonesia and New Guinea
evidence. In: Ruddle, K. & T. Akimichi, eds.
Maritime Institutions in the Western Pacific.
Senri Ethnological Studies 17. Osaka: National
Museum of Ethnology.
RANDALL, J.E., G.R. ALLEN & R.C. STEENE. (1990).
Fishes of the Great Barrier Reef and Coral Sea.
Honolulu. University of Hawaii Press.
RUDDLE, K., E. HVIDING. & R.E. JOHANNES. (1992).
Marine resources management in the context
of customary tenure. Marine Resource
Economics Vol. 7: 249–273.
SPC Traditional Marine Resource Management and Knowledge Information Bulletin #11 – September 1999
RUTTAN, L.M. (1998). Closing the commons: cooperation for gain or restraint? Human
Ecology 26(1): 43–65.
S UDEKUM , A.E., J.D. P ARRISH , R.L. R ADTKE & S.
RALSTON. (1991). Life History and Ecology of
Large Jacks in Undisturbed, Shallow, Oceanic
Communities. Fishery Bulletin, US. Vol 89:
493–513.
THRESHER, R.E. & J.S. GUNN. (1986). Comparative
analysis of visual census techniques for highly
mobile, reef-associated piscivores (Caran-
25
gidae). Environmental Biology of Fishes. Vol
17(2): 93–116.
VEITAYAKI, J. (1997). Traditional marine resource
management practices used in the pacific
Islands: an agenda for change. Ocean and
Coastal management 37(1): 123–136.
W ESTERNHAGEN , H. (1974). Observations on the
natural spawning of Alectis indicus (Ruppell)
and Caranx ignobilis (Forsk.) (Carangidae).
Journal of Fish Biology. Vol. 6: 513–516.
New publications
Traditional marine resource management and knowledge
Community-level sea use management in the Grenada beach seine net
fishery: current practices and management recommendations.
by J.A. Finlay,
M.Sc.Thesis, University of the West Indies, Cave Hill, Barbados (1996).
This study identified and characterised a community level, self-regulating Territorial Use Rights in
Fisheries (TURF) system within the Grenada
beach-seine fishery. The TURF system serves as a
useful example of an alternative management system and involves 41 large beach seines and 289
fishers, operating at 97 shallow water fishing sites
(hauls) along the coasts of the islands of Grenada
and Grenadines. Although all hauls are open to all
seine nets, nets are observed to cluster in nine fairly discrete zones along the coast and tend not to
move far from the fishers’ places of residence.
The beach-seine fishery targets a multi-species
stock of mainly juvenile coastal pelagics. Two
species of Carangidae (big eye scad, called jacks,
Selar crumenopthalmus; and red tail scad called
‘round robins’, Decapturus tabs) account for 90 per
cent of catches. The remaining catch comprises
juvenile oceanic pelagics (6%) and other smallsized coastal pelagics (4%). Although species vary
in abundance seasonally, a fairly constant overall
abundance of seine fish maintains a year round
fishery with relatively constant fishing effort.
Ten traditional rules are identified within the
TURF system and these effectively allocate fishing
opportunity to beach-seine nets through the recognition of temporary exclusive ownership at hauls.
A number of conflicts are presently threatening the
TURF system. These include intra-fishery conflicts
between beach-seine fishers and relate to non compliance with TURF rules; inter fishery conflicts,
between seine fishers and other fishers generated
by competition for sea space (by anchoring boats,
motorised trolling boats, motorised fishing boats
landing catches ashore); and competition for the
target fish by ringnets operating adjacent to the
haul; extra fishery conflicts, between beach seine
fishers and mainly tourist-related activities which
are now being viewed as having greater economic
importance. Beach-seine TURF fishers view these
conflict situations as devaluing their traditional
property rights.
Suggestions for solutions from both seine fishers
and their competitors for sea use, indicate that
government should initiate urgently some form of
localised management. Seine fishers felt strongly
that legalisation of the TURF rules would serve to
legitimise and enhance their property rights. Both
seine fishers and their competitors for sea use felt
that a rationalisation of entitlements, expectations
and obligations within the context of a co-management approach was needed.
Contributed by:
Dr. Robin Mahon, 48 Sunset Crest, St. James,
Barbados. E-mail: rmahon@caribsurf.com