By 1980 the American nuclear industry, along with its nuclear knowledge infrastructure, was in decline due to growing public concerns for nuclear safety, nuclear weapons proliferation, economic uncertainty, mediocre early reactor performance and waste disposal. Recent renewed interest in expansion of nuclear power production is now being accompanied by rising market demand for nuclear professionals and increasing awareness among policymakers of the need to improve the nuclear knowledge infrastructure to support industry growth. Expanding deployment of nuclear power in the USA will necessarily require reversal of the declines in all the elements of the nuclear knowledge base.

The behavioral-ecological traits of pelagic marine fish larvae diagnostic of specific ecological roles are discussed. Topics considered include: parental effects (egg size, yolk quantity, spawn distribution); swimming behavior; feeding and searching behavior; prey size relation- ships; and abundance and distribution of prey. Pertinent literature is reviewed and unpublished data on the larvae of northern anchovy, Engraulis mordax, and Pacific mackerel, Scomber japonicus, are pre- sented. These two species are used as examples of two distinctly dif- ferent ecological roles exhibited by marine fish larvae.

Laboratory measurements indicate that schooling begins in larval northern anchovy when they are between l l and 12 mm standard length and is well established when they reach 13-15 mm. The onset of schooling closely parallels an increase in patchiness of larvae in the sea, and it begins during the period that larvae form a duplex retina and undergo major changes in their respiratory and locomotor systems. RESUMEN Observaciones sobre larvas de Engraulis mordax en el laboratorio indican que la formacion de cardumenes se inicia cuando las larvas alcanzan de 11 a 12 mm de longitud normalizada, y el cardumen ya esta bien definido en larvas de 13 a 15 mm de longitud. La formacion del cardumen se produce coincidiendo bastante con el period0 en que la distribution de las larvas en el mar muestra un incremento en la formacibn de agrupaciones, y ademis se inicia cuando la larva adquiere una doble retina y experimenta cambios notables en sus sistemas respiratorio y natatorio. INTRODUCTION Except for a portion of the larval stage, all life activities of northern anchovy-feeding, avoiding predators, migrating, and reproducing-are conducted within schools. Thus the time at which schooling begins is an important event in the life history of anchovy, for it identifies the first time that they may be able to profit from schooling. In the larval stage the most important benefit of schooling may be a reduction in predation and cannibalism, although facilitation of the search for food and timing of vertical migration might be additional benefits. For example, fry of the freshwater fish Gobiomorus dormitor have a 30 percent chance of being eaten when alone, whereas one individual in a school has a chance of less than 1.5 percent (McKaye et al. 1979). Thus, identifying the larval size at which schooling begins is necessary to properly understand the effects of predation, cannibalism, and food abundance on size-specific mortality rates. [Manuscnp received December 4. 1981 I 246 Long Island Universlb Southhampton. New York 11968 Our objective was to identify the period when schooling begins, by observing the behavior of northern anchovy larvae reared in the laboratory. The ontogeny of schooling behavior has been studied in detail in Menidia menidia by Shaw (1960 and 1961) and Williams and Shaw (1971), in Atherina mochon by Jome-Safriel and Shaw (1966) and Williams (1976), and in various cichlids by Dambach (1963). The focus of these past studies was on describing the development of this behavior, the effects of isolation, and the behavioral mechanisms underlying school development. Our focus was more limited, as we wished simply to determine the minimum larval size or age at which schooling might be expected to begin in the sea. Thus larvae were maintained in large communal rearing containers rather than in isolation, and behavioral measurements were designed to identify the onset of schooling but not to understand the underlying behavioral mechanisms.

_ Thc visual feeding threshold and action spectrum of IO15-mm northern anchovy larvae were determined, using as criterion the incidence of two or more rotifers in the guts of 50% of larvae tested under different spectral irradiances. The threshold sensitivity to broadband blue-green light is at 0.15 pWcrn-’= 0.61 lux. The action spectrum shows a maximum in the green wavelengths around 530 nm: the weighted threshold irradiances at various wavelength bands converge at a mean value of 0.14 pWcm-* The visual abilities of the anchovy allow them to feed at a depth of 74 m at noon on clear days, and at the surface during twilight and bright nights, and appear to be well adapted to the anchovy’s habitat in turbid, greenish coastal waters. Comparisons with younger anchovy larvae show that changes in visual function accompany changes in eye and retinal morphology, specifically the recruitment of rods. Moreover, 10-15-mm anchovy larvae can feed to a limited extent (10%) in the dark when food densities are high (20-40 rotiferdml). In March, which is the peak spawning season of the anchovy in southern California and Baja California, the 10-15-mm larvae have 13 hours each day to feed. eff. RESUMEN Se determina en larvas de Engruulis mordux (anchoa) de 10-15 mm. de longitud, el umbral de visibilidad para capturar alimento y la acci6n del espectro, usando como criterio la incidencia de dos o mas rotiferos en el tub0 digestivo del 50% de las larvas sometidas a estudio bajo diversas condiciones espectrales de radiacion. La sensibilidad del umbral en la amplia banda de la luz verde y azulada se encuentra a 10s 0.15 ~ W c m ~ = 0 . 6 1 bujias. La accion del espectro sefiala un maximo en las longitudes de onda de la banda verde, alrededor de 10s 530nm: la estimacion del umbral de la anchoa en bandas de distinta longitud de onda, converge en un valor medio de 0.14 pWcm-2. La anchoa tiene una habilidad visual que le permite capturar alimento a una profundidad de 74 m. a1 ’Pre\eni addre\\ S E A F D K Aquacullure Depanment. P 0 Box 256. Il~iili~ Cily. Phhppine\ 5901 ]Manuscript received April 26, I983 I mediodia en dias despejados, y en la superficie del mar a1 atardecer y en noches claras. A1 parecer las anchoas se adaptan bien a su habitat de aguas costeras verdosas y turbias. Comparaciones efectuadas con larvas de anchoa mas jovenes, indican que 10s cambios en la funci6n visual corresponden con las variaciones en la morfologia del ojo y de la retina, especialmente con la incorporacion de 10s bastones. Las larvas de anchoa de 10-15 mm. de longitud, pueden alimentarse, con ciertas limitaciones (10%) en la oscuridad, cuando la densidad del alimento en el habitat es elevado (20-40 rotiferos por ml.). En el Sur de California y Baja California, el maxim0 de puesta de la anchoa se produce en Marzo, y entonces las larvas de 10-15 mm. de longitud disponen de 13 hords diarias de luz, que es suficiente para capturar el alimento. INTRODUCTION In this paper we determine the visual threshold and action spectrum of 10-15-mm northern anchovy (Engruulis mordux) larvae, and consider some implications of their feeding ecology. Studies on morphology and behavior have shown the importance of vision to anchovy (O’Connell 1963; Loukashkin and Grant 1965; Schwassmann 1965; Hunter 1972). At hatching, anchovy larvae are nearly transparent and have neither functional eyes nor jaws. O’Connell (1981) found that the oculomotor muscles differentiate at 3.5 mm SL; the photopic system is functional when feeding starts at 4 mm; and an area temporalis is present at 5 mm. The lens retractor muscle appears at 7 mm, and the rods at 10 mm. These developments indicate that an early feeding anchovy larva has a well-defined visual axis, good eye mobility, and binocular vision. The ability to accommodate to greater distances increases the perceptive field for feeding, and the recruitment of rods with commensurate increase in visual sensitivity increases the time that perception of food (and predators) is possible. Thus IO-mm anchovy can be expected to be more visually adept than 6-mm larvae. The ontogeny and maturation of sensory and locomotor systems and behavior patterns are important elements in larval survival (Hunter 1976a, 1977, 1981). The visual threshold and action spectrum were de-

Bathymetry is a key element in the demography of Dover sole Microstomus pacificus. Juvenile Dover soles begin life in relatively shallow water on the continental shelf but move gradually down the continental slope into deeper water as they grow and reach sexual maturity. Large and old individuals may eventually reach depths greater than 1,500 m. The ontogenetic migration of Dover soles into deeper water gives rise to a general pattern of increasing fish size with water depth, but there are differences between sexes and locations. To describe these patterns in greater detail, we computed probability distributions for depth, length, and sex using research trawl data for Dover sole taken off central California and Oregon. Our results indicate that females live at greater depths than do males and that abundance of large Dover soles may be lower at depths where lowest oxygen concentrations occur. In addition, major differences in bathymetric demography exist between study sites due, we suggest, to differences in bottom topography. We used one group of probability distributions and logbook data for Oregon and central California in a simple model to demonstrate how depth of fishing and bathymetric demography complicate interpretation of length composition and other data from a commercial fishery. Results from the model indicate that depth-stratified assessment models may be required to interpret fisheries data for Dover sole and other species, like shortspine thornyhead Sebastolobus alascanus and sablefish Anoplopoma fimbria that undergo ontogenetic migration from shallow into deep water.

A commercially made, pneumatically operated pump was modified for dispensing insect artificial diets into a variety of containers. Some accessories specifically designed for use with agar- based insect diets are described. When manually operated the modified machine can dispense a uniform preset amount of diet into 4 test tubes at once at a rate of 100 tubes per minute. The machine can dispense from 4 to 50 ml of diet per cycle into other types of containers. Keywords: insectsartificial dietsdispenserillustrations

Thc visual feeding threshold and action spectrum of IO- 15-mm northern anchovy larvae were determined, using as criterion the incidence of two or more rotifers in the guts of 50% of larvae tested under different spectral irradiances. The threshold sensitivity to broadband blue-green light is at 0.15 pWcrn-’= 0.61 lux. The action spectrum shows a maximum in the green wavelengths around 530 nm: the weighted threshold irradiances at various wavelength bands converge at a mean value of 0.14 pWcm-* The visual abilities of the anchovy allow them to feed at a depth of 74 m at noon on clear days, and at the surface during twilight and bright nights, and appear to be well adapted to the anchovy’s habitat in turbid, greenish coastal waters. Comparisons with younger anchovy larvae show that changes in visual function accompany changes in eye and retinal morphology, specifically the recruitment of rods. Moreover, 10-15-mm anchovy larvae can feed to a limited extent (10%) in the dark when food densities are high (20-40 rotiferdml). In March, which is the peak spawning season of the anchovy in southern California and Baja California, the 10-15-mm larvae have 13 hours each day to feed. eff.

The density threshold for the onset of filter feeding in adult anchovy (Engraulis mordux) fell between 5 and 18 Arfemia salina naupligl, between 1 and 2 an- chovy eggdl , and between 15 1 and 328 Gymnodinium splendens cellsll. Observations of the incidence of fil- tering schools indicate that behavioral changes oc- curred at lower food densities. These observations also indicate that the intensity of filtering by anchovy schools may be a function of the biomass of filterable foods. Filtering persisted long after the density in the tank was reduced below that required for the initiation of filter feeding. RESUMEN La densidad minima para el comienzo de alimenta- ci6n por filtrado en adultos de Engraulis mordux dis- minuy6 entre 5 y 1% nauplios de Artemia salina por litro, entre 1 y 2 huevos de anchoveta por litro, y entre 151 y 328 cilulas de Gymnodinium splendens por litro. Observaciones sobre la incidencia de car- dumenes en proceso de actividad filtradora indican que se producen cambios en su comportamiento cuando el alimento aparece en densidades bajas. Estas observaciones tambikn indican que la intensidad de filtrado en 10s cardumenes de anchoveta puede estar determinada por la biomasa de alimento filtrable. Este tip0 de alimentaci6n aun persisti6 mucho despuks de que disminuyera la densidad en el tanque, alcanzando valores por debajo del umbral que marca el comienzo del proceso de alimentacion por filtrado.

Abstract Photorepair of UV‐B lesions occurred in embryonic northern anchovy larvae. The photoreactive fluence rate required to fully activate photorepair mechanisms was less than 10% of that available from the sun on a clear day in March (33N). Even with UV–B enhancement from ozone depletion, sufficient photoreactive fluence exists in the sea to ensure maximal photorepair of UV damage in anchovy larvae.

Abstract— Northern anchovy larvae were exposed to various amounts of both natural (global solar) and artificial (sunlamps filtered by cellulose triacetate) UV‐B energy over a 12‐day period. Dosage was determined on the basis of a weighting function for biological effectiveness. The action spectrum on which this biologically effective dose for anchovy mortality is based was developed using broad‐band spectroscopy. These experiments indicated that biologically adverse conditions exist near the sea surface. Larvae exposed in shallow containers to global solar UV for 12 days suffered significant UV mortality from February to October. Larvae surviving all solar and artificial UV doses were smaller than those not exposed to UV. [Lowest dosage = 398 J·m−2(bio.eff.).] Fifty percent of the larvae survived a cumulative dose of 605 J·m−2(bio.eff.), or 50 J·m−2·day−1(bio.eff). Dose reciprocity did not hold; when a similar cumulative dose was given in the first 4 days of 12, there were about one‐half as many survivors.One meter below the sea surface the daily solar UV dose (corrected for average cloud cover) is equal to or greater than 50 J·m−2(bio.eff.) (33°N, 118°W), the daily equivalent of the LD50, for 7 months of the year (March to September) at moderate chlorophyll‐a concentrations (0.5 mg‐m−3). These data suggest that larvae in the sea may be subject to some UV‐B stress at present. Owing to their seasonality of spawning and vertical distribution, anchovy populations may not be gravely endangered by moderate degree of stratospheric ozone diminution.

The early life history of Pacific mackerel, Scomberjaponicus, is described from laboratory-rearing studies and examination ofstomach contents ofsea-caught larvae. At 19° C mackerel eggs hatched in 56 hours, larvae were 3.1 mmstandardlength with a dry weightof0.04 mg ofwhich 50% was yolk. First feeding occurred 46 hours afterhatching; all larvae fed by 60 hours (age 2.5 days). Larvae were then 3.6 mm long with fully pigmented eyes and 10% of the yolk remaining. Starvation was irreversible if larvae were notfed before age 4.5 days. Metamorphosis (15 mmstandard length) occurred in 24 days at 16.8° C to 16 days at 22.1° C. Larvae 3-5 days old consumed 87% oftheir body weight per day and had a mean gross growth efficiency in dry weight of33%. Oxygen consumption was 6.11-'1 0.. per milligram dry weight per hour at 18° C and 11.4 p.l O~ per milligram dry weight per hour at 22 C. Swimming speeds ranged from 1.3 standard lengths per second for first-feeding larvae to 3.8 standard lengths per second for fish at metamorphosis. Fifty percent of the larvae were able to capture a prey whenthe width of the prey was 85% ofthe width ofthe mouth and 95% were able to do BOwhen the prey was 57% ofthe width ofthe mouth. Cannibalism was common in rearing groups; at 8 mm standard length, 50% of the larvae became capable of feeding on other fish larvae and cannibalism ceased when schooling com­ menced. Chieffood items ofsea-caughtlarvae were stages of copepods; maximumfood width increased rapidly with larval length and was equivalent to the maximum mouth width. Mean prey width was 38% ofmouth width. The larger organisms, constituting halfofthe prey eaten, accounted for 85-90% of the total volume of food ea~n. The development and distribution ofeggs and lar­ vae ofthe western Pacific population ofthe Pacific mackerel, Scomberjaponicus, has been described (Kramer 1960; Kramer and Smith 1970), but little information exists on growth, behavior, and physiology of the larval stages. Incubation times and other data are known for the Japanese popula­ tion of S.japonicus (Watanabe 1970). This paper provides some of the information needed to characterize the early life history of Pacific mack­ erel, including incubation times, yolk absorption, onset of feeding, vulnerability to starvation, swimming and feeding behavior, food ration, and oxygen consumption.

Abstract— Anchovy and mackerel eggs and yolk‐sac larvae were exposed to UV radiation in the bioactive band of wavelengths between 280 and 320 nm. the UV‐B region of the spectrum. Irradiation levels were based upon predicted UV‐B increases that would result from anthropogenic diminution of Earth's protective ozone shell. Dose‐response relationships for mortality and histological and morphological effects were determined for two different spectral energy compositions, using FS‐40 sunlamps and two filter combinations. Anchovy were more sensitive than mackerel to UV‐B. Data for anchovy were analyzed in terms of DNA‐effective doses, i.e. the integrated spectral thence (in J/m2/nm) with the energy at each nm weighted by its effectiveness relative to the Setlow generalized DNA action spectrum. Fifty per cent of anchovy survived a cumulative DNA effective dose of 1150J'm‐2 over a 4‐day period. In the surviving larvae. irradiation induced lesions in the brain and eye. caused marked dispersion of pigment within melanophores and retarded growth and development. At the lowest dosage used. 760 (J. m‐2)DNA, growth was retarded and brain lesions occurred in anchovy. Calculations of Smith and Baker (in this issue) indicate that in clear ocean water a significant incidence of lesions and retardation of growth in anchovy could occur at the surface at a 25%, reduction in ozone and down to 3.5 m at a 50% reduction. Eggs and larvae of anchovy occur at these depths.

The object of this review was to provide an understanding of the behavioral factors that determine whether or not larvae of the northern anchovy, f ipauhs mordax Girard, will starve. The swimming and feeding behavior of anchovy larvae is described and related to temperature, developmental stage, characteristics of the food and food distribution and to other variables. The principal sources of this information were published and unpublished observations of larvae cultured in the laboratory. INTRODUCTION Predation and starvation are considered by most specialists to be the dominant factors influencing survival of larval fishes. In the Southwest Fisheries Center (SWFC) of the National Marine Fisheries Service (NMFS), we have concentrated on the problem of starvation although some studies have been made on predation (Lillelund and Lasker, 1971; Theilacker and Lasker, 1974). The object of this review is to provide an understanding of the behavioral factors that determine whether or not an anchovy larva will starve. The swimming and feeding behavior of anchovy larvae is described and related to temperature, developmental stage, characteristics of the food and food distribution, and to other variables. The principal source of this information is observations of larvae cultured in the laboratory. Experimental design will not be given. where information is already published, but a brief description will be given for new data. For information on laboratory culture of northern anchovy, the reader should consult Lasker, Feder, and Theilacker (1970), Theilacker and McMaster (1971) , and Hunter (1976). EGGS AND YOLK-SAC LARVAE In southern California the major spawnings of anchovy occur in February and March but some spawning continues throughout the year (Lasker and Smith, 1977). During the height of the spawning season, temperatures in the upper 10 m vary from about 13" to 16" C. Eggs are transparent oblate spheroids about 1.34 mm long and 0.66 mm wide, and are neutrally buoyant. Larvae hatch in 89.8 hrs at 13" C and at 59.8 hrs at 16" C (Zweifel and Lasker, 1976). The larvae at hatching averages 2.86 2 0.028 mm standard length and weigh 0.0246 2 0.0014 mg dry weight, of which 53% is yolk. At hatching larvae are nearly transparent, have no functional eye or jaw but olfactory and lateral line organs are developed (C. O'Connell, National Marine Fisheries Service, La Jolla, pers. comm.). After hatching, larvae are inactive; over 90% of the time they float motionless in the water usually with head directed downward. About once a minute they execute a burst of intense swimming, lasting about 1 to 2 sec. These bursts of activity may have a respiratory function as they occur regularly and increase in frequency when oxygen concentrations are below saturation (Hunter, 1972). The larvae have no functional gill filaments at this time so movement could serve to increase the transport of gases across the integument. Owing to their inactivity, small size, and lack of a functional visual system, it appears anchovies in the yolk sac stage must be extremely vulnerable to predation. Indeed, Lillelund and Lasker (1971) showed that the copepod Labidocera tzkpinosa were 60% successful in capturing 1 day old yolk-sac larvae whereas success decreased to 11 % for larvae that had begun to feed (4 days old). These authors also point out that success of capture seems to be correlated with the degree of activity of the larvae. Development of a functional visual system and jaw, and nearly complete absorption of yolk coincides with the onset of feeding and a major increase in locomotor activity. Typically, larvae average about 4.0 mm standard length (SL) at this time but means vary from 3.8 to 4.4 mm depending on average egg size. The mean size of anchovy eggs varies seasonally, with larger eggs being more common in the winter (February and March) and small eggs occurring more frequently during the summer months (Lenarz and Hunter, MS). This seasonal variation would be expected to produce a seasonal trend in average larval size at first feeding. A larger size at first feeding would probably be of greater adaptive advantage in the winter months when cold temperatures produce slower growth rates than in the summer when it is more rapid. SWIMMING BEHAVIOR OF POST YOLK-SAC LARVAE Knowledge of swimming abilities of larval fish is essential for an understanding of survival. For example, the cruising speed of a larva will determine the frequency it will encounter prey, and determines in part whether or not a larva will remain in areas REPORTS VOLUME XIX, 1 JULY 1975 TO 30 JUNE 1976 139 where prey are concentrated; also swimming accounts for most of the larva's energy expenditure. Burst speed capabilities determine in part the ability of a larva to avoid predation and plankton nets. From the onset of feeding through adulthood, anchovy swimming consists of two types: continuously propagated caudally directed waves typical of most fishes that swim by caudal propulsion, and a series of bursts of motion consisting of a single tail beat followed by a rest or a glide. The beat and glide mode of swimming or intermittent swimming is used for cruising, whereas the other mode, continuous swimming, is used for high speed bursts. Nearly all swimming is intermittent in adult anchovy and in larvae after feeding begins. Continuous swimming in larvae is used for occasional bursts lasting usually less than a second and with a frequency of about 12 burstslhour, less than 1% of the time devoted to swimming (Hunter, 1972). Vlymen (1974) developed a model for energy expenditure of larval anchovy during intermittent swimming from theoretical considerations and analysis of cine photographs. He found that larvae of 14 mm standard length expended 5 x 10-3cal/hr, and had a metabolic efficiency of 25%. This efficiency was high when compared to values obtained for larger fish and led Vlymen to speculate that this may have given anchovy larvae a competitive advantage over the Pacific sardine, Sadnops caerdea. It should be mentioned parenthetically, that although intermittent swimming may be more efficient, it is slow, and consequently the volume of water that can be searched for prey is less than that of a fish that swims in the continuous mode. Swimming speeds of larval anchovy have been measured in various ways: plotting positions from cine photographs; counting tail beat frequencies; and visual approximation of larval position against a grid. For a temperature of 17" to 18" C the mean swimming speed of a 5 mm larval anchovy was 3.0 mmlsec using the photographic technique, 4.1 mmlsec using counts of tail beat frequency (Hunter, 1972); and 4.5 mml sec when a grid was used (Hunter, unpublished data). These data set a range of cruising speeds of 0.6 to 0.9 body lengthslsec for anchovy larvae at 17" to 18°C. Maximum bursts of speed, on the other hand, appeared to be about 15 body lengthslsec but swimming in this case is continuous rather than intermittent and can be sustained for only brief periods (Hunter, 1972). Although it is clear that larval size is a significant determinant of speed, the methods used to measure speeds of larvae, however, are not sufficiently accurate to establish the correct length coefficient. To do this larvae must be subject to known water speeds. Consequently, to adjust for size, the convention of dividing speed by length has been followed (Bainbridge, 1958), but it should be recognized that the true coefficient is probably less than unity (Brett, 1965). In addition to size, cruising speeds are influenced by many other variables, thus the above averages can be considered only to be general estimates. Temperature, age, food distribution, feeding activity, and condition of the larva all influence swimming speed. Speed declines in starving larvae until they remain almost motionless, floating head down in the water. Food distribution may have several effects on swimming speed. In dense patches of food ( Gymnodinium splendens) speed declines by about 60%. On the other hand, at any given density the larvae that are feeding at the highest rate are generally the ones that are swimming the fastest (Hunter and Thomas, 1974). This latter effect might be caused by the physical state of the larvae. For example, within a given food density the healthy and presumably better fed larvae are the most active and feed most often. REST AND INTERMITTENT SWIMMING (percent)

Incidental mortality of porpoise (Cetacea, Delphinidae) occurs in the tropical tuna seine fishery. Experiments were carried out in a crowding chamber to determine behavioral responses of trained and naive Hawaiian spinner porpoise (Stenella cf. S. longirostris) to barriers of purse-seine netting, monofilament webbing, polyvinyl sheeting, rows of floats, and openings of various dimensions in a net wall. The object of the experiments was to generate information to be used in development of rescue gear and methods for the fishery. Openings of less than 1.5 m in width and/or 1 m in depth markedly inhibited escape. Negative effect of a line of floats across an opening at the surface was pronounced. Barriers of visually and acoustically relatively transparent monofilament webbing and polyvinyl sheeting were not apparently detected by porpoise prior to physical contact. rescue gear are presented. Incidental mortality of porpoise occurs in the American purse-seine fishery for tropical tunas (Perrin, 1970). In 1970, the National Marine Fisheries Service began a program of research to develop improved gear and methods to reduce the porpoise mortality due to tuna seining. This paper reports the results of experiments on the responses to netting and other barriers by the Hawaiian spinner porpoise (Stenella cf. S. longirostris), a form closely related to one of the species involved in the tuna fishery.’ We studied the response of the spinner porpoise to barriers of net, transparent monofilament nylon webbing, transparent polyvinyl sheeting, rows of floats,