FAO Fisheries Circular No. 920 FIRM/C920
REVIEW OF THE STATE OF WORLD FISHERY RESOURCES: MARINE FISHERIES
Marine Resources Service,
Fishery Resources Division,
FAO, Rome, Italy
2. LANTERNFISHES: A POTENTIAL FISHERY IN THE NORTHERN ARABIAN SEA?
The possibility that lanternfishes could support globally significant fisheries has been known since the 1960s. There are many species of lanternfishes in the Myctophidae family and together with Gonostomatidae species they account for around 75% of the total global catch of small mesopelagic oceanic fishes. Lanterfishes (or myctophids) typically have a maximum size of 7-8 cm (standard length, SL), with individuals in this size range weighing 2-6 g. While distribution is worldwide, production appears to be highest in tropical and sub-tropical areas. There are few examples of commercial fisheries targeting lanternfishes. A Soviet fishery for Diaphus coeruleus and Gymnoscopelus nicholski (species which were considered edible) in the Southwest Indian Ocean and Southern Atlantic began in 1977, and catches by former USSR countries reached 51 680 t in 1992, after which the fishery ceased. Despite this, the Commission for Conservation of Antarctic Marine Living Resources (CCAMLR) still permits a TAC for this fishery of 200 000 t in its convention area. An industrial purse seine fishery for Lampanyctodes hectoris in South African waters closed in the mid-1980s due to processing difficulties caused by the high oil content of the fish.
In the late 1970s and early 1980s researchers investigated the feasibility of the development of a commercial fishery for mesopelagic fishes in the northern Arabian Sea. These studies indicated that such a fishery may be commercially feasible, especially for Benthosema pterotum in the Gulf of Oman region. After a long period of high expectations, 1996 has seen the start of a commercial fishery for these mesopelagic fishes in the Gulf of Oman.
LIFE HISTORY AND BIOLOGY
The highest density of myctophids in the region was in the Gulf of Oman, with B. pterotum being the only species. Other species (Benthosema fibulatum, Diaphus spp., Myctophum spinosum or Symbolophorus evermanni) were occasionally numerically more common than B. pterotum in the Gulf of Aden. Along the southern Omani and northeastern Somali coasts, B. fibulatum dominated trawl collections and acoustic surveys records. In the eastern and northeastern Arabian Sea, Diaphus arabicus and D. thiollieri were common. Along the coast of Pakistan myctophid concentrations consist almost exclusively of B. pterotum with densities decreasing to the west.
Temporal and spatial distribution
The fish ascend into shallow (<100 m) or surface waters (<5 m) at night and return to deep water (typically 250 to >1000 m) during the day (i.e. a 24 hour, or diel, migration). However, certain life history stages often migrate little, if at all, and in some populations sizable fractions of the population normally associated with vertical movement may not do so every day or at all. Echo sounder records show that many myctophid species aggregate in compact layers, especially during the daytime when they are relatively quiescent. These aggregates are the primary component of the acoustically dense "Deep Scattering Layers". Their densities, which correspond to concentrations of five-ten individuals per cubic meter and trawl catches of up to 10 to 20 t/hr have led to commercial fishery feasibility trials.
Concentrations of B. pterotum usually divide into two deep daytime layers (an upper 'D1' layer and a deeper 'D2' layer) and two shallower night layers (an upper 'N1' layer and deeper 'N2' layer). The D1 layer is centered at about 150 m (range 130 to 200 m) and is the most densely concentrated of the diel layers; abundances can exceed 8 individuals per m3. The D2 layer is more diffuse and is centered at 250 m (range 150 to 350 m). A descent of individuals from D1 to D2 beginning a few hours after sunrise and a return movement to the D1 layer commencing two to three hours before sunset has been noted, although on other occasions only a single daytime layer has been observed. Daytime layers are usually deeper (250 to 350 m, occasionally 300 to 500 m) in offshore waters of the Arabian Sea. Vertical migration at night usually starts about one hour prior to sunset and is essentially complete between one-half to one hour after sunset. The N1 zone ranges from near the surface to 100 m and is centered between 20 and 50 m, whereas the more diffuse N2 zone ranges between 220 and 300 m. In addition to being the densest layer, the N1 layer, where distinct from the N2, essentially contains only B. pterotum. Other layers include other myctophid species, other fishes, shrimps, euphausiids and salps.
Observations based on a fifteen-month acoustic data set indicates a regular four-fold variation in stock size of B. pterotum in the Gulf of Oman. The variations follow the summer-winter monsoon pattern, with highest stock estimates during May and June (beginning of summer monsoon) and lowest in November (beginning of winter monsoon). This cyclic pattern could be due to variation in productivity caused by upwelling. Data from different periods between 1975 and 1984 also show a pattern of variability which is correlated with the monsoon pattern.
Life history parameters
Most of the Arabian Sea myctophid species are small, typically less than 50 mm SL at maximum size. In the Gulf of Oman, B. pterotum rarely exceed 40 mm and most individuals are less than 35 mm. Data suggests a life span of less than one and that two generations may be produced within a year. Age at maximum size is seven months and ten months for the Gulfs of Aden and Oman respectively. Sexual maturity in females is at relatively small sizes (~ 40 mm SL) and at an age four to five months.
Reproductive output of B. pterotum is low: females between 27 and 52 mm SL collected off Mozambique produced between 200 to 3 000 eggs. Lower fecundity has been reported from the Gulf of Oman (210 - 1 334 eggs in individuals 27 to 48 mm SL). Studies of the seasonality of reproduction in the Arabian Sea found actively spawning individuals during all seasons, but there appeared to be increased spawning during monsoon transition periods (March-June and September-November).
Several studies have assayed the chemical composition of myctophids and other mesopelagic fishes. These studies show that from a nutritional standpoint myctophids are high in proteins, variable but lower in lipids and uniformly low in carbohydrates. A number of studies have evaluated the lipid content of vertically migrating myctophids and found that they include both triglycerides, believed to serve primarily as an energy store, and wax esters, mainly used for buoyancy.
STOCK ASSESSMENT OF TARGET SPECIES
The stocks of myctophids have been assessed by both scientific trawl and acoustic surveys. Neither of these approaches is considered entirely accurate alone (it is suspected that the small trawls are too easily avoided, that escapement is high from the large trawls and that precision of acoustic surveys is low due to uncertainty over the actual backscattering cross-section per unit weight of fish sonified). Therefore acoustic surveys are typically combined with one of the trawl approaches.
In the late 1970s and early 1980s a number of Norwegian researchers reported on catch rates of mesopelagic fishes along the coasts of Somalia, Oman, Iran, Pakistan, and also in the Gulfs of Oman and Aden. As these concentrations consist essentially of only B. pterotum, it is easier to ascertain catch rates, stock sizes and life history parameters affecting stock size. The catch rates reported were highly variable among and within cruise periods and regions, even over small areas such as in the Gulf of Oman proper.
Norwegian data in the Gulf of Oman from the mid-1970s found average catch rates from krill and Harstad trawls (250 m2 and 400 m2 opening cross-sections respectively) to range between 0.05 and 2.5 t/hr. Larger commercial pelagic trawls (650-700 m2) had mean catch rates of 3.8-5.0 t/hr, with occasional massive catches of 20 to 100 t/hr. More recent trawl survey data from the Gulf of Oman reflect these earlier findings for this region. The largest catches were consistently in the D1 layer (from 4.5 to 80 t/hr, average 37 t/hr) while D2 catches were below 1 t/hr and the N1 layer yielded an average of 10 t/yr. Acoustic survey data collected at the same time found only a few regions of high density concentrated in the westernmost part of the Gulf.
Total stock estimates in the Gulf of Oman range from 1.7 to 20 million tonnes. The difference in estimates may be due to the seasonal variation in abundance and/or differences in methodology. While studies that focused on the commercial aspects of mesopelagic fishes in the northern Arabian Sea-Gulf of Oman regions have discussed variations in hourly catch rate and standing stocks, few have presented estimates on potential yields. One report for the eastern Gulf of Oman estimated a potential annual yield for B. pterotum of 2.3 million tonnes which was believed to be a conservative figure, but this estimate does not imply 'availability' of this amount of stock to commercially viable fisheries. Given the annual variability of survey data coupled with the short lifespan of B. pterotum and the patchy distribution of densities of this species regionally, it is unlikely that a single MSY figure will be useful for management of the stock over an extended period.
Variability in catch data among and between cruises, regions and seasons has been observed by all researchers. In addition to natural population variability, two other sources of variation in catch rates arise from the different aims of scientific surveys and the different gear used. For instance, the mid-70s cruises mainly focused on assessing the species composition of the acoustic scattering layers and calibrating the acoustic data, while late 1970s and 1980s cruises were attempts to employ commercial-sized fishing gear and obtain high catch rates. Different gears were used for each of the different studies, leading to differences in net escapement. The smallest mesh sizes in the various commercial trawls employed was about 9 mm but significant levels of net escapement of smaller myctophids occurred in scientific trawls of 1.6 mm vs. 4 mm meshes. Fish of different lengths will have different vulnerability to gear depending on the size structure of the population and gear used.
POTENTIAL EFFECTS OF A FISHERY
No studies have examined the potential effects of a commercial myctophid fishery on other mesopelagic fish populations, and few have dealt with other organisms that ecologically interact with myctophids. Since fishing effort and ecological impacts are linked (in a complicated and unknown way) the absence of any data on future probable effort levels confounds the evaluation of potential impacts. However, a commercial fishery for B. pterotum focused on the high densities in the D1 layer in the Gulf of Oman will have little direct effect on other mesopelagic fishes, since the D1 layer is essentially monospecific. Extension of the range of fishery eastward along the coast of Pakistan, into the Gulf of Aden, or south into the Arabian Sea, will alter this scenario as mesopelagic diversity in general increases in all of these areas. Likewise, trawling in the deeper D2 zone or the night-time zones will capture non-target species to a highly variable degree.
The potential effects of the removal of a sizable fraction of myctophids from the Gulf of Oman by fisheries operations also raises the issue of the effect on their zooplankton prey and on their competitor species (salps, other fishes and carnivorous crustaceans). However, the important stocks of tunas, mackerels and billfishes that inhabit the northern Arabian Sea and Gulf of Oman region are unlikely to be too affected because myctophids appear to form an insignificant part of their diets. However, it is possible that in the Gulf of Oman and other northern Arabian Sea regions, where the oxygen content of the water has limited or eliminated the presence of other small fish species, B. pterotum may be a more significant component of the diet in at least scombrid fishes. While many cephalopods are piscivorous, especially in larger species, or at larger sizes, there are virtually no data on the nature of feeding in larger cephalopods in the available literature. At least 20 species of cephalopods have distributions that range into the northern Arabian Sea-Gulf of Oman region and, of these, at least nine species exhibit depth ranges that would bring them into contact with mesopelagic fishes.
A number of cetacean species prey on myctophids and among the non-mesopelagic animals that may be affected by a commercial fishery for myctophids, these are likely to be the most influenced. Stomachs from representatives of three cetacean families in the Pacific Ocean, the Delphinidae (five species), Phocoenidae (1 species) and Physeteridae (1 species) showed myctophid remains. The delphinids especially appear to be dependent on myctophids. Two of the delphinid genera and three species (Delphinus delphis, Stenella attenuata and S. longirostris) and the physeterid Kogia simus have been recorded from the Gulf of Oman and northwestern Arabian Sea. However, among ocean-dwelling cetaceans, the level of dependence is variable.
It is clear that B. pterotum should be the focus of exploratory commercial fishing and continued scientific research efforts. Other species that could be considered for study prior to commercial exploitation are B. fibulatum and D. arabicus. The zone for concentration of fishery efforts should be D1. While all reports indicate that the stocks of mesopelagic fishes in general and B. pterotum in particular are large and that it is unlikely that overfishing of the entire resource could occur, there are several reasons for caution in the development and management of such a fishery.
For instance, since B. pterotum spawns during the late afternoon-evening hours while migrating to the surface, an intensive fishing of the D1 layer during the day is going to constantly remove reproductive individuals from the population before they can spawn. Concerns over the maintenance of recruitment may have effects on strategies for sustainable exploitation. There is also large uncertainty attached to stock estimates as mentioned above. This uncertainty arises from a number of factors such as high spatial and temporal variability in catch rates, question marks over stock identity (and associated regional differences in lifespan and growth) and the strength of seasonal effects on growth and abundance. Regarding stock identity, current growth rate estimates and data on reproductive patterns suggest that separate aggregations of B. pterotum may exist in the Gulf of Oman, Gulf of Aden and perhaps in the northern Arabian Sea. The wide range in densities by region means only certain areas will support commercial fisheries. It is possible that locally-intensive fishing could deplete myctophids in limited areas and if these different regions are populated by different sub-populations, the spawning sub-populations could become over-fished.
It is probable that for this species, that stock estimates (and management decisions) must be done at least annually. If seasonal variability is cyclic, tied to the monsoon seasons, then semi-annual evaluations of stock are advisable. An additional reason for caution is that the potential effects of a fishery on the ecosystem are not adequately understood. Although there is agreement as to the importance of myctophids in the cycling of energy in oceanic ecosystems, there have been few studies evaluating myctophid interactions with higher trophic levels. At least among the cetaceans, correlations have been observed between loss of favoured prey items and abandonment of a region.
Using the most recent estimates for catch rates and biomass, certain simple calculations can be made. The most recent estimates of maximum average catch rates in the D1 layer is about 37 t/hr and potential annual maximum sustainable yield is 2.3 million tonnes. If one assumes that fishing is limited to the D1 layer only, beginning one hour after sunrise and ending one hour before sunset (when the fishes have settled into the D1 layer) for a 10-hour fishing day, approximately 370 t/day would be landed, over twice the vessel�s processing capacity. If the average N1 layer catch is included (10 t/hr) beginning one hour after sunset and ending one hour before sunrise, daily (20 hours fishing) the catch rises to 470 t. If 470 t/day was realistic, and if maximum effort (365 fishing days) were possible, total annual tonnage landed by the one vessel would be 171 550 t, approximately 7.5% of the MSY estimate for the eastern part of the Gulf of Oman. A more realistic assessment might be based on other data in which a total of 17 trawls were fished in eight days for a total of 900 t, or roughly 112 t/day. Based on this figure, trawling every day of the year would produce about 41 000 t/vessel. A more realistic 250 fishing days would net about 28 000 t.
These assumptions are obviously extremely simplistic. Variations in catches are extreme, often few hundred kilograms to a few tonnes even in the D1 layer. Time of year is significant, and catches may fall to negligible levels over wide regions. It was found that fish densities in the D2 layer (average 3 t/hr) were too low to be commercially viable. This further reinforces the possibility that fishery operations will be conducted intensively in localized regions, particularly the western end of the Gulf of Oman.