Estimating the Number of Fish in Atlantic Bluefin Tuna (Thunnus
Total Page:16
File Type:pdf, Size:1020Kb
420 Abstract–Aerial photographic assess- Estimating the number of fi sh in Atlantic ment is a promising technique that could be structured to yield a fi shery- bluefi n tuna (Thunnus thynnus thynnus) schools independent index of abundance for Atlantic bluefi n tuna, Thunnus thyn- using models derived from captive nus thynnus (ABT). The accuracy of this school observations approach may be increased by incor- porating the relationship between the surface characteristics of a school and Brian Hanrahan the total number of individuals. Our Francis Juanes objective was to develop models to facil- itate the estimation of number of fi sh in Department of Natural Resources Conservation ABT schools from aerial photographs. University of Massachusetts Video cameras were used to observe Amherst, Massachusetts 01003-4210 74 incidences of schooling for 50 cap- E-mail address (for F. Juanes, contact author): [email protected] tive ABT approximately one meter in length. Relationships between the sur- face characteristics of ABT schools and the number of fi sh in the school were explored by using least-squares regres- sion. The schools ranged in number from The bluefi n tuna (Thunnus thynnus) is has been based upon landings data and 2 to 45 individuals. A weighted regres- distributed worldwide in temperate and abundance indices (Scott et al., 1993). sion model incorporating the number subtropical seas. It has a limited dis- Use of ABT landings data to generate of fi sh in the school at the surface tribution in the southern hemisphere. an abundance index may lead to bias as the independent variable and the Endothermy by means of vascular heat due to variability in effort, improve- number of fi sh in the remaining por- exchangers allows bluefi n tuna to inhabit ments in fi shing technology (Lo et 2 tion of the school yielded an r of 0.74. a wide thermal niche and therefore wide al., 1992), and variability in annual A second weighted multiple-regression geographic and depth ranges (Carey and geographic distribution linked to prey model incorporating the number of Teal, 1969; Carey and Lawson, 1973). In distribution.1 These characteristics of fi sh in the school at the surface and in the second depth interval (0–25% the western Atlantic Ocean, the Atlantic the northwest Atlantic bluefi n tuna school depth below surface layer) of the bluefi n tuna (ABT), Thunnus thynnus fi shery, in conjunction with popula- school as independent variables, and thynnus, is distributed from Labrador tion-level behavioral characteristics ob- the number of fi sh in the remaining to Brazil, including the Gulf of Mexico served for similar tuna species, suggest portion of the school as the dependent and Caribbean Sea. Adult ABT occur that the use of catch-per-unit-of-effort variable, with 1/variance as the weight, throughout the entire range, but smaller (CPUE) data to evaluate tuna popula- achieved an r2 of 0.70. A third model bluefi n tuna (less than 45 kg) are not tion trends could lead to inaccurate es- using the length and width of the sur- observed frequently above the latitude timates (Clark and Mangel, 1979). The face layer of the school as the indepen- of Cape Cod, Massachusetts. The Atlan- accuracy of CPUE-based assessments dent variables and the number of fi sh tic bluefi n tuna is epipelagic and usu- in estimating the abundance of bluefi n in the school as the dependent vari- able had an r2 of 0.86. One data point ally oceanic but appears near the coast tuna in the northwest Atlantic remains from a wild school is currently avail- seasonally (Squire, 1962; Collette and controversial (Clay, 1991; Suzuki and able to verify model predictions. This Nauen, 1983) to feed on concentrated Ishizuka, 1991; Safi na, 1993). An exten- school of 125 individuals is well outside assemblages of prey. Adult ABT may sive discussion of the issues involved in the range of school sizes used to con- attain a length of four meters and a Atlantic bluefi n tuna assessment can be struct the model (2–45 individuals), yet body mass of 680 kg. Large medium found in the National Research Council differs from model predictions by only (178–195 cm FL) and giant (>195 cm report by Magnuson et al. (1994). 7%. FL) bluefi n tuna are targeted by com- Recent investigations have focused We believe that these models have the mercial purse-seine, long line, and hook- on the feasibility of using aerial pho- potential to improve an abundance index and-line fi sheries (Mather, 1974; Figley, tographic assessment of large medium based on aerial photographs by estimat- ing the number of individuals in wild 1984). A recreational hook-and-line fi sh- and giant bluefi n tuna in New England ABT schools from surface characteris- ery (Mather, 1974; Figley, 1984) targets waters and in the Straits of Florida tics observed in aerial photographs. all sizes of bluefi n tuna as they appear (Lutcavage and Kraus, 1995; Lutcav- along the east coast of the United age et al., 1997) as an alternative fi sh- States and Canada from June to Octo- ery-independent method of obtaining ber (Mather, 1962). indices of abundance. The aerial survey The combination of changes in the spatial distribution over time and as- sociated uncertainty regarding the in- 1 Chase, B. C. 1995. Preliminary report of dependence of eastern and western At- the Massachusetts bluefi n tuna investiga- tion: the diet of bluefi n (Thunnus thynnus) lantic stocks makes the estimation of off the coast of Massachusetts. Massa- Manuscript accepted 26 January 2001. ABT stock size particularly problemat- chusetts Division of Marine Fisheries. Fish. Bull. 99:420–431 (2001). ic. The stock assessment for this species Salem, MA 01947, 39 p. Hanrahan and Juanes: Estimating the school size of Thunnus thynnus thynnus 421 method has been used to determine the relative abun- (e.g. total count, biomass) of schools (Partridge et al., 1983; dance for other pelagic fi sheries worldwide, including En- Lutcavage and Kraus, 1995). In addition, the behavioral graulis mordax (Lo et al., 1992), Engraulis mordax, Sar- and environmental factors that may infl uence tuna school da chiliensis, Trachurus symmetricus, etc. (Squire, 1972), structure and dynamics remain poorly described (Mather, Trachurus declivis, Katsuwonus pelamis, Arripis trutta, 1962; Clark and Mangel, 1979; Partridge et al., 1983). Thunnus maccoyii (Williams, 1981), Mugil spp. (Scott et Our study presents a functional relationship between al., 1989), and Squire (1993) has reported aerial survey the surface characteristics of and the total number of indi- data for Thunnus thynnus orientalis and other species. viduals in ABT schools. We analyzed video-taped footage Abundance estimates derived from an aerial assessment of 74 incidences of schooling in a group of captive ABT to are based on biomass or number of individuals per unit of quantify the relationship between the number of fi sh vis- area. ible at the surface and the total number of individuals in Lutcavage and Kraus (1995) concluded that the aerial the school (NFS), the relationships between school dimen- method could provide area-specifi c minimum abundance sions (e.g. length, width) and NFS, and to explore the effect and distribution data for large medium and giant Atlan- of environmental conditions within the net-pen enclosure tic bluefi n tuna under good viewing conditions. However, on school size and dimensions. We also analyzed the verti- many diffi culties associated with aerial photographic as- cal distribution of individuals within schools across school sessment of ABT remain to be resolved. Sea state, light- size, and propose a mechanistic explanation for the limited ing conditions, and turbidity all play an important role size of the two-dimensional schools observed by Partridge in the ability to detect and produce useful photographs et al. (1983) and Lutcavage and Kraus (1995). We then ap- of schools (Lutcavage and Kraus, 1995). Rough seas, sun ply the predictions from one of the resultant models to the glare, and high turbidity may all result in reduced detec- single open-ocean school size estimate available. tion of schools, limiting the days on which this type of sur- vey method is effective. Visual counts of individuals at the surface derived from aerial photographs are diffi cult to in- Methods terpret without a verifi cation count and information on behavioral factors such as surfacing frequency (Lo et al., Field methods 1992) and the proportion of the school visible at the sur- face (Lutcavage and Kraus, 1995). In addition, variability We employed a 30.5-m diameter, 15.3-m deep, cylindrical in population movements and distribution could lead to an fl oating net-pen enclosure (Fig. 1) to hold the tuna used in inaccurate abundance estimate if an intensive, spatially our study. This enclosure is similar to those used in tuna expansive sampling scheme is not employed. research and culture operations around the world. Its low We propose a technique to address the problem of es- cost, large internal volume (11,128.5 m3), and its resiliency timating the number of fi sh in a school (NFS) from the to dynamic and often damaging effects of the offshore envi- surface characteristics of a school. If the relationship be- ronment make this enclosure the most appropriate type tween the surface structure or the surface number of fi sh for observing the behavior of large pelagic fi sh in captivity. and number fi sh in total school was known, school sur- The enclosure proved to be very resilient to the damaging face counts from aerial photographs or visual observa- effects of a close pass of a hurricane and a tropical storm. tions could be adjusted to include an estimate of total A white, one-inch, straight-hung mesh net constituted the NFS, facilitating an improvement of area-specifi c mini- vertical walls and bottom of the enclosure.