Appendix E. Alternative Viewpoint for Near-Term U.S. GLOBEC Myctophid Studies in the Arabian Sea

Laurence P. Madin, Stephen M. Bollens, James E. Craddock

As detailed in the Fish Working Group report and scientific review of fish biology sections, myctophids are very important in the Arabian Sea ecosystem. Their densities in the northern part of the Sea are higher than any observed elsewhere (Gjosaeter, 1984). They undergo diel migration down into the oxygen minimum zone, potentially accelerating vertical fluxes. Their biomass peaks in winter and spring, possibly in response to monsoon effects. The significance of the fish populations to U.S. JGOFS objectives seems clear: "If these (very high) fish biomass numbers are accurate, then any attempt to balance a carbon budget cannot fail to include a high level of effort at determining the biomass of mesopelagic fishes, their daily rations, and their impact on zooplankton" (Peterson, 1991 p. 105). The population dynamics of these organisms, and their likely responses to monsoonal forcing of production cycles make their biology equally central to any proposed U.S. GLOBEC efforts. The question is how best to sample the fish, and whether that sampling method is affordable and logistically compatible with other goals of U.S. GLOBEC or U.S. JGOFS. We agree with the desirability of the sampling equipment and approach recommended by the Fish Working Group. However, if use of chartered fishing vessels is not an economic possibility, we believe that information on midwater fish and macrozooplankton populations crucial to U.S. GLOBEC and U.S. JGOFS goals can be obtained using smaller and more manageable MOCNESS trawl and acoustic systems during process-oriented research cruises aboard a UNOLS vessel. We presume there will be coordination of planning and overlap of investigators between the U.S. JGOFS and U.S. GLOBEC efforts, and we want to argue for the importance and feasibility, in both programs, of sampling midwater fish and macrozooplankton with research-scale gear.

Larger nets do not necessarily provide better estimates of population size. Potter et al. (1990) compared catches of 3-5 cm young cod collected with a 10 m2 MOCNESS (3 mm mesh, 2 kt tow speed) and the 104 m2 International Young Gadoid Pelagic Trawl (105-5 mm graded mesh, 3.5 kt tow speed). They found that abundance estimates were at least as high with the MOCNESS as with the IYGPT, with very little difference in size distribution of the fish caught.
The results of Potter et al. were a primary reason why the 10 m2 MOCNESS was chosen as the best gear to sample juvenile cod and haddock in the U.S. GLOBEC Georges Bank/Northwest Atlantic Field Program. These trawls have been proposed for use in both broad scale surveys and process-oriented sampling.
Myctophids are in general less active fish (Barham, 1971; Craddock, pers. obs.) than the cod fished by Potter et al. (1990), and probably less able to avoid capture in nets. MOCNESS trawls have been and are now being used to sample midwater fish in parts of the Atlantic that are far more oligotrophic than the Arabian Sea.
Different size trawls were also used in collections of myctophids of the Benguela region (Hulley, 1986). Catch rates from three nets, an RMT-2 (2 m2), RMT-8 (8 m2) and MT-1600 (ca. 300 m2), were inversely proportional to net size, decreasing from 0.2474 to 0.0565 fish per 1000 m3. Stock estimates based on catches of the three nets were very similar, except for the mean size of the fish caught. There was little overlap in sizes caught by the smallest and largest nets, so that stock estimates based on their catches probably represent different parts of the population.
The commercial trawls used by Gjosaeter (1984) in the Arabian Sea caught far more fish than researchers could handle. The 250 m2 net had mean catch rates of 200 kg h-1 in 1979 and 1000 kg h-1 in 1981. The 750 m2 trawl used in 1983 had a mean catch rate of 4700 kg h-1. We don't need (or want) that many fish. A first-order scaling of the mouth areas back to the 20 m2 MOCNESS ("MOC-20") we propose using (12.5 and 37.5 times smaller mouth area) indicates it might catch between 16 and 125 kg h-1. For research purposes, commercial gear is not only unnecessary, it is undesirable.
Kinzer et al. (in press) used three much smaller nets, an 8 m2 IKMT, a 4 m2 Jungfisch-Trawl, and a 1 m2 BIOMOC, to sample myctophids in the eastern Arabian Sea. The IKMT caught tens to hundreds of the commoner species per haul, sufficient to estimate vertical distribution and abundance, as well as provide material for stomach analysis.
Commercial trawls are open nets that cannot provide accurate data on depth distribution. In contrast, the MOC-20 can fish 6 discrete depths down to 2000 m, and provide continuous environmental data (CTD, O2, etc.) in real time. The 3.0 mm mesh means the MOC-20 will sample post-larval fish better, providing better size data to model population dynamics and growth rates.
An important sampling goal will be to catch fish in good condition for shipboard studies of metabolism, diet, and bioluminescence. We expect the MOC-20 to catch fish in better condition, both for identification and experimental use, because of the relatively soft mesh and the smaller number of fish in the cod end. We also anticipate fitting thermally-protecting cod ends to the nets.

These observations lead us to conclude that a 20 m2 MOCNESS should be at least as accurate as a commercial trawl for estimating myctophid abundances, and much more effective for sampling vertical distribution and migration, and retrieving experimental specimens. The MOC-20 can be easily fished from a UNOLS vessel, along with other sampling work. It is towed from the stern, but launched and recovered over the side, so that the net frame is stabilized against the hull of the ship as it enters and leaves the water. Despite its size, the trawl is well suited to work in rough weather. The frequent and intensive sampling schedule recommended by the Working Group is clearly desirable for midwater fish (as it would be for zooplankton). But if it is not fiscally possible, we suggest that accurate, depth specific sampling with the MOC-20 and acoustic equipment should be carried out along with other pelagic process-oriented work aboard the UNOLS ship. We hope that U.S. GLOBEC will not dismiss or defer the study of myctophids in the Arabian Sea for simple logistic reasons. If the scientific questions surrounding the population biology and ecology of these fishes are of sufficient interest and importance -- as we agree they are -- then technology and expertise are currently available to sample these organisms in a quantitative and cost-effective manner.

References

Barham, E.G. 1971. Deep-sea Fishes: Lethargy and vertical orientation. pp. 100-118 in Farquhar, G.B. (Ed.), Proc. Int. Symp. Biol. Soundscattering Layers. Washington D.C. Maury Center for Ocean Science, Rept. Mc-005.

Gjosaeter, J. 1984. Mesopelagic fish, a large potential resource in the Arabian Sea. Deep-Sea Res., 31, 1019-1035.

Hulley, P.A. 1986. Lanternfishes of the Southern Benguela Region. Part 1. Faunal complexity and distribution. Ann. S. Afr. Mus., 97, 227-249.

Kinzer, J., R. Bšttger-Schnack, and K. Schulz. In press. Aspects of horizontal distribution and diet of myctophid fishes in the Arabian Sea with reference to the deep water oxygen deficiency. Deep-Sea Res.

Peterson, W.T. 1991. Zooplankton and nekton in the Arabian Sea. pp. 95-106 in U.S. JGOFS Arabian Sea Process Study, U.S. JGOFS Planning Report No. 13, Woods Hole Oceanographic Institution, Woods Hole, MA.

Potter, D.C., R.G. Lough, I.R. Perry, and J.D. Neilson. 1990. Comparison of the MOCNESS and IYGPT pelagic samplers for the capture of 0-group cod (Gadus morhua) on Georges Bank. J. Cons. int. Explor. Mer, 46, 121-128.