Objective: Jointly with nominated MAFF staff (Mr Sam Mario), to plan and execute a survey of Gracilaria spp. (and any other seaweed species with known commercial potential) in the Suva/Rewa/Tailevu regions and interview villagers, in order to locate any significant populations and to gain an impression of the distribution, abundance, habitat-type and seasonality of the various species found. Information to be gathered from seaweed fishers and other sources about harvest and marketing methods, and prices paid.
Visits were made by boat (18' Yamaha skiff) to various places in Suva Harbour and Kubuna Waters, to locate any significant beds of economic seaweeds. Subsequently, a visit was made to a place near Serua Island after a report that seaweeds could be found growing there. Notes were made at each locality to describe bottom substrate type and water depth. At some localities measurements were made of water temperature and salinity using a thermometer and hand-held refractometer. A qualitative assessment of seaweed abundance was also made. Quantitative assessment using quadrats was not carried out owing to the high sampling effort required to get meaningful data and the limited boat time available to cover such a wide area. Small samples of plants were collected at each site and preserved in 5% formalin in seawater for subsequent identification. If sufficient seaweed was found, fresh plants were gathered in sacks and later dried to provide samples for phycocolloid analysis.
The sites visited are marked on the maps shown in Figs. 1,2 and 3.
Suva Harbour was surveyed on 24 January 1994 (low tide at 0543, high tide at 1210), over an area from Laucala Island to Muaivuso. Kubuna Waters from Kaba Peninsula to just south of Ucunivanua (Verata) was surveyed on 1 – 2 February 1995 and on 14 February 1995. Serua was visited on 19 September 1996.
The findings of the survey at each site visited are summarised below. The numberss 1,2,3 correspond to those marked on the locality maps in Figs. 1,2 and 3. Names given to the seaweeds are as determined during the laboratory examination described in Section 3.0, and as determined by the various experts consulted.
(1) Laucala Island
Substrate is mainly silty sand, with some rock outcrops above low-tide mark and mangroves at high-tide mark. Patchy turtle-grass beds begin from the low-tide mark and extend into the high subtidal zone. At the time the site was visited (8:00 hr) these beds were under about 1 m of water on a rising tide. Attached to or entangled with the turtlegrasses were small (>100 mm) seaweed plants identified as Gracilaria maramae and Hypnea pannosa. G. maramae was also found attached by holdfasts to buried pebbles or shells. Plants of Acanthophora spicifera were attached to rocks in the intertidal zone. None of the species found was very abundant (at a guess, probably less than 50 g wet wt per m2). However the site had been previously visited during June and July 1994, at which time Gracilaria plants in particular were much larger (up to 500 mm in length) and bushier, as well as more abundant.
(2) Corpus Cristi
The substrate is sand with scattered patches of “kuku” (date mussels), on a windward shore facing the prevailing southeasterlies. G. maramae plants up to 500 mm in length, attached by holdfasts, were found along the edge of the intertidal mudflat (near the navigational marks) in the high subtidal zone where the bottom slopes away into the shipping channel. Unattached plants were also found entangled in kuku beds and washed up into the intertidal zone. Small plants of Hypnea pannosa, Acanthophora spicifera and a previously unidentified species subsequently identified as Gracilaria edulis (found at low water mark), were also found here. None of these seaweeds were very abundant. Although G. maramae plants were quite large, they were found scattered as individual plants and did not form any dense beds at the time the site was visited.
(3) Nasese (subtidal mudflats on either side of Leveti creek mouth)
This area is large, and extends over 500 m from the shore to the edge of the shipping channel. There are in places dense beds of turtlegrasses on a soft silty-sand substrate, which at 1030 hr was under approximately 1 m of water. The area is in the lee of Suva Point so is protected from any choppy seas caused by prevailing winds, yet tidal currents still provide a high level of water mixing. G. maramae plants were large and very abundant in this area, and formed dense beds of unattached plants entangled with and often smothering the turtlegrasses. Individual plants were up to 1 m long, and stock density was estimated to be in the neighbourhood of 1000 g wet weight per m2. Isolated clumps of Gracilaria edulis were found beneath the layer of G. maramae plants, and could be distinguished by their tightly-knit mossy or turf-like appearance. Three sacks were easily filled with G. maramae for phycocolloid analysis. This site represented the largest Gracilaria resource encountered during the entire survey.
Smaller amounts of scattered plants attached by holdfasts can almost always be found in sand and rubble areas nearer to shore. G. maramae is most often found as single attached or unattached plants in tidal pools, while G. edulis forms a close-knit turf of small interlocked plants on the raised hard rock surfaces.
During the 4 years since this visit in January 1995, no comparable amounts of Gracilaria have since been observed in this seagrass area. It appears to have been a one-off “bloom” of Gracilaria, perhaps due to a particularly favourable combination of environmental conditions at that time; a similar “bloom” of Gracilaria was observed along the Vatuwaqa shoreline in late 1998, when several shipping containers could easily have been filled.
At Nasese the inshore populations of Gracilaria and other species are more persistent, though at times one or other species can disappear almost completely only to reappear a month or so later. Sometimes large unattached plants of G. maramae (but never G. edulis) can be found trapped in soft sediment by the river mouth and persist for some time, indicating either a tolerance to lower salinity levels than G. edulis, or a morphology more amenable to such entrapment.
Fig. 1. Map of Suva harbour, showing survey sites
(4) Nawanada backreef area (opposite Nasese)
At 1100 hr, this area of backreef was under 2.5 m of water. Bottom substrate is mainly dead corals, with some live hard and soft corals. Sea urchins (Echinodermata) were abundant, and there was an almost complete absence of palatable macroalgae.
(5) Nasese/Nasova (near Ratu Sukuna Road junction)
This area consists of rocky bottom interspersed with sandy intertidal pools and patches of turtlegrass, which merge into a bare sand bottom as water deepens into the channel. Exposed at low tide, the bottom was under about 1.5 m of water at 1130 hr. Plants of Hypnea pannosa, Gracilaria maramae and Gracilaria edulis were abundant, but were fairly small. Hypnea in particular was very stunted, being usually >50 mm in length and often covered with filamentous epiphytes. Small Acanthophora spicifera, Gelidiella acerosa and Laurencia sp. plants were also found attached to rocks. G. maramae and G. edulis can here be found growing beside each other in identical environments while displaying major differences in growth habit. G. maramae has a long, sparsely branched and quite thick thallus that gives it a characteristic “spaghetti-noodles” appearance. It is usually found growing as discrete plants. G. edulis has a very short, much-branched thallus that is thinner in diameter, and it grows in dense clumps that form a moss-like turf (see Plate 5). The appearance of these two contrasting forms in the same environment suggests that they may be different species rather than just examples of morphological plasticity within a single species.
(6) Salia Reef (navigation mark opposite Nasova Police College)
Though almost exposed at low tide, this reef was under almost 2 m of water when visited at 1200 hr. The bottom is mainly dead coral rubble, with live soft corals. In a roughly 200 × 500 m area around the navigation mark, there was an abundance of Acanthophora spicifera plants attached by holdfasts to rubble and soft corals. They formed dense clumps of plants up to 150 mm long. Further out, the Acanthophora thins out and is replaced by brown algae with some small epiphytic Hypnea plants. Two sacks were filled with Acanthophora for phycocolloid analysis.
This is a backreef area of mixed coral rubble and sand. The dominant seaweed was scattered clumps of a kelp-like brown alga with branches that float up into the water column. Tiny fragments of Hypnea were found, but they were rare at this locality. A single cystocarpic Gracilaria plant was found attached to a coral rock. It is very fleshy, with thallus about 5 mm in diameter. This plant has been tentatively identified as Gracilaria euchumoides.
(8) Draunibota Island
The area south of Draunibota Island and along the mangrove-bound coast toward Waiqanaki consists of sandy substrate with extensive turtlegrass beds. At 1400 hr, the bottom was under 1.0 – 1.5 m of water. Although it is a large area superficially similar to the Nasese turtlegrass beds, seaweeds were not abundant. Isolated small Gracilaria maramae and Hypnea pannosa plants were found entangled with turtlegrass, but there was not enough to be worth collecting.
(9) Udaya Point/Draunibota Bay
The water in Draunibota Bay is fairly deep, is ringed by mangroves and encloses the Veisari rivermouth. Substrate at the mangrove edge is soft mud, with no evidence of any seaweeds or turtlegrasses. It is believed that a combination of low salinity, deep water and poor water mixing may prevent seaweeds from thriving here.
(10) Nukumarorika (Mosquito) Island
Tiny fragments of Gracilaria maramae plants were found in sand beside a rocky promontory at the south end of the popular picnic beach, but there were no significant amounts of seaweed found at this locality.
There is a large area of intertidal mudflat along the foreshore of Lami township, with sandy bottom at the eastern (Suvavou) end merging into coral rubble and soft sand/mud at the western end. A few small G. maramae plants were found attached by holdfasts in the high subtidal zone. Turtlegrass beds in some tidal pools close to the beach by Lami Town held a few small Hypnea fragments. The western (Nukumarorika Island) end has a lot of tidal pools and shallow subtidal areas with sandy bottom, but no noteworthy seaweeds were found. The dominant seaweeds in this area were Gelidiella acerosa and Enteromorpha, plus a lot of filamentous and sheet-forming microalgae. The area has the appearance of a relatively eutrophied and polluted environment, and the sand is covered with a thin layer of fine brown silt from land runoff.
(12) Kaba Peninsula
Five main localities were surveyed at low tide along the seaward side of Kaba Peninsula. These localities, Saulailai, Sakotuvu, Nautodua, Wainivai and Navau, correspond to landmarks known to local residents. The whole stretch of coast from Kaba Point down to Kaleli Settlement was fairly uniform, having a sandy substrate in both the intertidal and subtidal zones, with occasional rocky outcrops in the intertidal. The subtidal area down to 1.5 m depth was dominated by turtlegrass and extensive beds of the edible green seaweed Caulerpa racemosa (nama). The horizontal stolons of Caulerpa plants were a conspicuous feature of the seabed, and other seaweeds (mainly Hypnea pannosa) were often found attached or tangled with them. Water temperatures were around 28 – 30°C, and salinity was 28 – 30 ‰. Acanthophora spicifera plants were common on rocky areas of the intertidal, usually on the headlands along this coast and particularly within 1 km of Kaba Point itself. In between the rocky headlands were intertidal sandy pools, often near mangroves and areas with freshwater runoff. These pools were lined with a mix of species but dominated by at least two species of Laurencia, epiphytized Hypnea plants, G. maramae and G. edulis. There was sufficient G. maramae present to enable collection for phycocolloid analysis, but other species were too small (>50 mm) and dirty to warrant collection. In the subtidal zone, larger (150 mm) and cleaner Hypnea plants could be found tangled in Caulerpa stolons, however the distribution of Hypnea was very patchy. Local residents advise that Hypnea and Gracilaria plants become larger and much more abundant from May through to October.
(13) Backreef opposite Navau
The backreef areas of places like Namotu Reef have a similar sandy substrate at similar depths to the Kaba Peninsula shoreline, but are almost bare of seaweeds. Only a little Caulerpa was found in this area. It is possible that grazing pressures are greater in backreef areas compared with the shoreline.
(14) Qalotu Reef
This sandy-bottomed backreef area was similar to site m above, and there was an abundance of beche-de-mer. A single drift Kappaphycus (formerly Eucheuma) plant was found that weighed about 20 kg. Presumably it was a survivor from the now-abandoned seaweed farms that used to be established northeast of this reef.
(15) Nukusasa sandbar
This bar of bare sand is just exposed at low tide. Water temperature was 30°C and salinity was 18‰. Along the water line were unattached fragments of G. maramae, Acanthophora spicifera and filamentous green seaweeds such as Enteromorpha sp. At the northern end there was a small patch of scattered G. maramae plants attached by holdfasts to shell buried in the sand. There were similar patches at the southern end, and on the southwestern side some turtlegrass beds with small Hypnea plants attached. Local residents report this to be a good area for “lumi” when it is in season (May to October). It is possible that higher temperatures and lower salinities in this river-influenced locality during the ‘hot’ or ‘rainy’ season (November to April) may inhibit seaweed growth compared with the ‘dry’ season.
(16) Tokatoka ni Kubu Reef
The rocky weather side of this reef had a lot of Acanthophora spicifera plants growing on it. The sandy lee side has scattered plants of G. maramae and H. pannosa.
(17) Cakalevu Reef
At low tide pools of water are formed amongst sand and coral rubble in the leeward backreef area on the southern edge of this reef. These pools were lined with many small (<100 mm) Hypnea plants, clean and healthy-looking and attached by holdfasts. This has the appearance of a good site for collection of Hypnea when it is in season. Scattered G. sp. plants were also found in these pools. Acanthophora spicifera plants were attached to rocks in deeper water at the reef's edge. The reef was exposed when visited at 1500 hr on 2 February 1995 (low tide was at 1422 hr). Water temperature was 31°C and salinity was 29‰. Hypnea plants were collected for phycocolloid analysis.
(18) Viwa Island
By cruising along the south coast of Viwa Island and looking down into the water along the fringing rocky shelf, it could be seen that Acanthophora spicifera plants were quite common down to about 1 m depth (time: 1515 hr). No other macrophytes were seen, however some women were met fishing who reported that Hypnea (lumicevata) can be found there from June onwards each year.
Fig. 2. (a) Map of Kubuna Waters, showing survey sites
Fig. 2. (b) Map of Kubuna Waters, showing survey sites
(19) Cost between Kubuna Island and Sebi Point
This coast has a shoreline of soapstone cliffs with a wave-cut platform at sea level. No seaweeds were found here.
(20) Coast from Matasawalevu Landing to Vatoa
This coast is formed by mangrove swamp extending inland from the water's edge, with a soft black mud substrate almost exposed at low tide. At Cautata the water temperature was 31°C and salinity was 18‰. No seaweed was found growing in situ. At Vatoa drift plants of G. maramae, Hypnea and Enteromorpha were entangled in mangrove roots, and it is presumed that these were washed from the adjacent Tokatoka ni Kubu Reef.
(21) Telau Island
At the western end of this island is a shallow area of sandy bottom under about 0.5 m of water at low tide. Scattered small Hypnea and G. maramae plants were found entangled with turtlegrasses and Caulerpa stolons. There was not enough to be worth harvesting at that time.
This coast appears fairly uniform from Ucunivanua down to Kumi, and is superficially similar to the seaward coast of Kaba Peninsula. Coarse sand forms tidal pools, with turtlegrass beds in the high subtidal and rocky substrate in deeper water (>0.5 m). Acanthophora was common on these deeper rocks. In the sandy pools were found G. maramae and small, epiphytized Hypnea plants. This coastline is reputed to be a good source of lumi for metropolitan markets, so should be worth a visit when in season.
At the rocky point south of Kumi were found many large (up to 200 mm) Acanthophora plants attached to rocks and purple soft corals in about 0.5 m depth of water. A sample was collected for phycocolloid analysis. The water temperature was 31°C and salinity was 25‰. No other seaweeds were found.
(24) Navulivatu Settlement
This area was visited after a report by Gerald Billings of Fiji Fisheries Division that some seaweeds of interest may be found growing in Serua. Upon making a visit, four species of Gracilaria, along with some Caulerpa plants, were found in an area of inter-tidal and high subtidal seagrass meadow on the shoreline near Navulivatu Settlement. The area is bounded on the landward side by mangroves, and the waters of the lagoon on the seaward side.
One species was obviously G. maramae South, while the others were subsequently identified as G. edulis (Gmelin) Silva and G. arcuata Zanardini v. snackeyi Weber van Bosse (H. Yamamoto, pers. comm.). The fourth species, distinctive in its bright emerald-green colour and arcuate branching pattern, could not be identified and is possibly a new species (H. Yamamoto pers. comm.).
Thalli of all four species were terete, but varied in diameter. G. edulis and G. sp. had the thinnest thalli, being about 1.5 mm in diameter, while G. maramae thalli were about 2 mm and G. arcuata var snackeyi up to 5 mm in diameter. G. edulis often has a turf-forming growth habit, profusely branched but with short internodal distances, forming dense and interlocked mats of short (<50 mm) plants attached to hard substrate. At Serua, longer unattached plants were found, that formed balls and relied upon entanglement in seagrasses for attachment and were opaque khaki brown in colour. G. maramae generally has long internodal distances and is not profusely branched, so often resembles spaghetti noodles. Plant length can be up to 500 mm though 200 mm is typical. Plants were found individually scattered on sandy substrates, attached by a holdfast to pieces of shell and were slightly translucent golden brown in colour. G. arcuata v. snackeyi typically forms short (<200 mm) but thick, heavy and very rigid plants that entangle in seagrasses and are a light golden brown, with a characteristic pattern of small lightly-coloured spots. G. sp. resembled the turf-forming G. edulis in its growth habit, but the plants were bright emerald green and some plants had a distinctive arcuate branching pattern.
Subsequent visits to Serua have shown that the population of seaweeds at the Navulivatu site is never large, and in fact plants may be completely absent during the hotter months of the year. If these plants were found to have any commercial value, an immediate problem will be finding enough seedstock upon which an aquaculture project could be based.
Figure 3. Map of Serua District, showing survey site.
The Fisheries Division of MAFF collects statistical data on seafood sold in metropolitan markets. Data for wet weight and price per kg of seaweeds on sale, collected on one Saturday each month since January 1992 at Suva market, has been extracted from this database and is presented in Figs. 4 and 5. Because of the wide range of species that they need to monitor, MAFF does not distinguish between Hypnea pannosa (lumicevata) and Gracilaria maramae (lumiwawa) when data is collected; these are reported together as a single category “lumiwawa”. This can mask any trends for these species separately, for example hardly any Hypnea (lumicevata) was seen on sale between January and March 1994 so the wet weight figures represent almost 100% Gracilaria (lumiwawa). At other times of the year, however, most of the lumi sold will be Hypnea. In order to discover any accurate trends for one or other species, special one-off surveys are required.
It can be seen from Fig. 4 that there has been a steady increase in the amount of edible seaweeds offered for sale in Suva Market since 1992. There are no clearly discernible seasonal trends in this data, however harvesters may well increase their catch of lumiwawa when lumicevata is absent so that, taken together, there is the appearance of a steady supply of lumi. Non-seasonal factors that limit harvesting, such as weather and tide conditions, may produce background “noise” that further masks any seasonal trends in the data. It is also possible that the quantities offered for sale are demand-driven rather than supply-driven, where the data simply reflects a steady demand for 30 – 40 kg per Saturday and more than this cannot be sold even when plants are abundant.
Fig.4. Amount of seaweed on sale
Fig. 5. Seaweed prices
Fig. 5 shows that the prices paid for seaweeds jumped up during the aftermath of Cyclone Kina (December 1992), and have not come down again since. There has been a three-fold increase in seaweed prices over the space of 18 months (between the end of 1992 and mid-1994) from F$0.90 – 1.00 per kg to F$3.33 – 4.00 per kg.
A simple questionnaire was designed (attached as Appendix 1) in order to find out the harvesting areas and seasonality of seaweed species sold in Suva and Nausori market. This information, collected in mid-January, was intended to assist the site visits by narrowing down the areas that need to be visited. It soon became clear, however, that January is during the off-season for lumi. On Saturday 28 January 1995, no Hypnea pannosa was on sale at all (it has only been seen infrequently since). There was some Gracilaria maramae ($2.00 per heap) which a vendor had collected near Verata. It was generally found in muddy places and near mangroves, according to the vendor. The main seaweed on sale was Caulerpa racemosa (six vendors, selling at $2.00 per heap) which came from the Yasawa Islands (Naviti Island and Navula Island are good sources) and from Vatuloa (near Navua). This was usually found on clean sandy bottom. One vendor had a pile of the green seaweed Enteromorpha sp. (lumivoso) for sale at $1.00 per heap, collected from Vatuloa. This species is very seldom sold, and its presence in the market probably reflected the lack of other edible seaweeds.
Vendors were unanimous in saying that the season for lumi (Hypnea pannosa and Gracilaria maramae) begins in May/June and ends in October/November. During this time, they said most of the lumi which appears in the Suva and Nausori markets is harvested from the Verata coast, or from places near Navua (such as Veivatuloa). Vendors did not think that there was any season as such for Caulerpa, and attributed any variations in supply to other factors such as bad weather or inappropriate tides.
This study commenced during a time that turned out to be the “off-season” for edible seaweeds. Even so, many of the sites visited had populations of small plants which will most likely become much larger and more abundant when in season. It is therefore possible to identify, even in the off-season, where the most abundant seaweed populations can be found. Particular mention must be made of the Nasese beachfront in Suva, and the shore along Vatuwaqa, where Gracilaria maramae can at times become very abundant even in the off-season.
Since there have not yet been any detailed studies made of population dynamics and environmental factors for Fiji's seaweeds, one can at this stage only form a rough impression of the type of area likely to be conducive to seaweed growth. Areas where the seaweeds of interest appeared to be doing well usually had a sandy substrate and a water depth from 0.1 to 1.0 m at low tide, with good water exchange yet sheltered from choppy seas that might wash plants away (Nasese fits this description). As a general rule, sites that lacked these seaweeds tended to be either too exposed to wave action, or if sheltered had silty/muddy substrates, poor water exchange or persistently brackish water (such as near mangrove swamps and rivermouths). Gracilaria tended to dominate in areas where there was likely to be moderate or periodical freshwater influence, while Hypnea and Acanthophora were more often found in places where freshwater influence was unlikely or infrequent. Hypnea was often associated with Caulerpa beds, probably due to the ability for drift plants to snag and bind onto the horizontal stolons. An apparent puzzle is the lack of palatable seaweeds in backreef areas opposite beaches where such seaweeds are present. It is possible that different grazing pressures may be an explanation for this.
While economic seaweeds are fairly easy to find, nowhere could they be described as being in industrial-scale quantities or in densities that would enable mechanised harvesting. Some areas (particularly Nasese and Verata) support artisanal fisheries (hand-gathering) for Hypnea and Gracilaria when in season, so could serve as sources of seedstock for establishment of seaweed farms. None, however, have seaweed populations to match the Gracilaria beds found in places like the Philippines, Chile or New Zealand's Manukau and Ohiwa Harbours. The lack of dense stands of seaweeds is apparently typical of many tropical areas, and may in part be due to high grazing pressures. The impression gained from the survey of Suva and Tailevu (though without the benefit of any quantitative assessment of abundance and distribution) is that industrial-scale quantities of local seaweeds for phycocolloid extraction could only come from aquaculture activity.