Diel feeding pattern and diet of rough scad Trachurus lathami Nichols , 1920 ( Carangidae ) from the Southwestern Atlantic

Diel changes in feeding activity and dietary composition of the rough scad Trachurus lathami were investigated through the analysis of stomach contents of 307 fish sampled over a 24-h period on the continental shelf off Ubatuba (23 35 ́S 45W). Stomach contents were analyzed by frequency of occurrence (O%), percent number (N%), percent mass (M%), and feeding index (FI% = O% * M%). Rough scad fed on mollusks (Gastropoda, Crepidula sp.), crustacean (Ostracoda, Copepoda, decapod larvae), chaetognat and teleostean fish. The main items were calanoid copepods (Eucalanus sp. and Centropages sp.). According to the analysis of the stomach fullness and prey digestion, T. lathami is both a diurnal and nocturnal feeder, showing some seasonal variation in feeding time. According to the Chronobiology concept, it was raised the hypothesis of circadian rhythm in feeding of this fish, probably synchronized by light/dark cycle.


Introduction
The critical and limiting resources for all animals are food and time.Food provides energy and nutrients for reproduction, and to develop, maintain, and protect the somatic body.Time is necessary for foraging, as well as for other activities, such as resting and reproduction (Wootton, 1990).According to chronobiologists, the majority of organisms present a circadian rhythm in their activities and this organization is the result of an evolutionary process that expresses a synchronicity among geophysical, environmental and biological events (Marques & Menna-Barreto, 2003).Among marine, estuarine, and freshwater fishes, there are several records of feeding activity with circadian rhythm (see Neilson & Perry, 1990;Boujard & Leatherland, 1992, Sánchez-Vázquez et al., 1997, Gerkema et al., 2000).
Studies on diel feeding in the natural environment are ba-sic tools to understand the autoecology of fish species.And studies on diet composition may contribute to clarify preypredator relationships, and to elaborate trophic models for the assessment of stocks (Christensen & Pauly, 1992).Most species of Trachurus are widely distributed in both temperate and tropical coastal waters.Fifteen species are recognized (Suda et al., 1995).Only Trachurus lathami Nichols 1920 occurs in the south-western Atlantic from the Gulf of Maine (United States of America) down to the north of Argentina.It is a demersal-pelagic species, with maximum size about 40 cm of length and 0.5 kg of mass (Menezes & Figueiredo, 1980), occurring in waters from 50 to 100m in depth, at temperatures between 15 o and 17 o C and salinity between 35.5 and 35.8 (Saccardo & Katsuragawa, 1995;Mello et al., 1992).
Trachurus lathami is an abundant species on the southeastern Brazilian coast and represents a potential fishing resource (Saccardo & Katsuragawa, 1995).
Considering that there are few studies on diet composition of T. lathami (Pedreira, 1997;Crispino, 1995) and there are no studies about diel feeding of this species, we proposed to identify its pattern of daily feeding activity, and to describe the variation of its diet composition during the course of the day.
Different water masses are present in this area during summer and winter.During summer, there is a two-layer vertical stratification due to the seasonal thermocline: the upper layer consists of warmer and less saline Coastal Water -CW (T=22-24 o C; S=35-35.4) and the lower of colder South Atlantic Central Water -SACW (T=14-15 o C; S=35.6-35.8).During the winter, only Coastal Water is present on the inner shelf (Castro & Miranda, 1998).

Materials and Methods
Sampling.Rough scad were sampled during the periods 22-24 July 1988 and 4-6 December 1988 on the continental shelf off Ubatuba, São Paulo State, Brazil (Fig. 1).On each cruise, samples were collected in two consecutive 24-hour periods at 3-hour intervals, through 30 min trawls.
From the total catch, a sample of at least 30 specimens were randomly separated for biological analysis, size considered suitable for this type of study (eg.Livingston, 1982); all available individuals were collected from the small samples.A total of 307 stomachs were considered in this study (Table 1).
The specimens were frozen, and the total length and total mass of each individual were measured in the laboratory.The stomachs were removed and preserved in 10% buffered formalin solution.
Regional samples of this species are deposited in Museu  Laboratory methods.A fullness level was assigned for each stomach: empty, half full and full (Hyslop, 1980).The stomach content was sorted taxonomically and the degree of digestion (1-fresh prey; 2-partially digested prey; 3-digested prey) was assigned for individual items, according to Soares & Apelbaum (1994).The state of digestion of the total stomach content was assigned based on the proportion of prey in a particular state in the content: fresh, more than 75% of the prey in a fresh state; partially digested, 25-75% of fresh content and digested, less than 25% of fresh prey.
Data analysis.Diel feeding activity was analyzed according to data on stomach's fullness levels (%), stomach index of fullness (BM% = mass of the stomach content X mass of the fish -1 X 100), state of digestion of the stomach content (%), and number and frequency of fresh prey.
The samples were joined into a single 24-hour cycle per sampling period; the midpoint of each hauling period was used as a reference point in time.Sampling times that were less than half an hour apart were brought together into a single sample, and the midpoints averaged (Table 1).Variations in the BM% and number of fresh prey between sampling periods of the day were tested by Kruskall-Wallis' nonparametric test variance analysis (Zar, 1996) to assess the influence of the sampling time on feeding activity.

Results
Oceanographic conditions.In this study, during the winter, water temperature ranged from 19 to 21 o C and salinity from 32 to 35, presenting a homogeneous profile throughout the day.In the summer, two water masses were present, one of them with temperatures between 21 and 23 o C and salinity between 35.3 and 35.9, and the other with temperatures between 15 and 19 o C and salinity between 35.5 and 36.The temperature profiles are shown in Fig. 2.
Diel feeding activity.The stomach's fullness levels and index of fullness showed the same diel pattern during the winter and summer (Figs.3-4).Full stomachs occurred in the afternoon and at night.Empty stomachs occurred only in the summer, in the early morning and at dawn.Data on index of fullness revealed a pattern of daytime feeding, attaining maximum values during the afternoon in the winter (Kruskall-Wallis test: c 2 =47,09; P<0.0001) and summer (Kruskall-Wallis test: c 2 =138.93;P=0).
The index of fullness presented the same tendency along the day, in winter and summer, when fishes were divided in adults (>115 mm) and juvenile (<115 mm), following the estimative of the size of first maturity (Saccardo & Katsuragawa, 1995) (Fig. 5).
The proportion of stomachs containing fresh prey was highest in the afternoon (15:33h) and at night (18:32h) in the winter.During the summer, stomachs with fresh or partially digested prey were highest at afternoon (15:20h, 17:23h) and at night (20:16h) just stomachs with partially digested prey occurred (Fig. 6).
Diet.The diet of T. lathami was composed of Mollusca, Crustacea, Chaetognatha, and Teleostei fish (eggs and larvae. Among Arthropoda there was a greater abundance of Crustacea, represented by Ostracoda and Copepoda (Tab.2).
The diet composition was similar over the feeding activity time (afternoon and night) during the winter and summer, and consisted mainly of calanoid copepods (Eucalanus sp. and Centropages sp.) and decapod larvae (Tab.3).
Two similar groups of length class were detected based on diet composition.Specimens of the classes of 50 to 90mm presented similar diet composition, mainly cumaceans, and the classes of 100 to 130 mm ingested essentially calanoid copepods (Tab.4).

Discussion
Trachurus lathami presented distinct phases of feeding and nonfeeding time during the 24-hour period, suggesting a circadian rhythm in feeding activity.Evening and nocturnal feeding activity were observed.We can say that even in the summer and in the winter, the feeding activity occurred, mainly, at afternoon and night phases.This similar result can be explained by the fact of the distinct points where the sample was collected (38m in winter and 72m in summer) were located in an intermediate area.This area presents different oceanographic conditions (Castro & Miranda, 1998), but this species in large schools can occupy all of this area, including the different points cited (Almeida, 2001;Saccardo & Katsuragawa, 1995).
The majority of the fish known as "horse-mackerel" (Trachurus spp.) present diurnal feeding habits (Shuntov, 1969;Pillar & Barange, 1998).Daily variation in feeding activity is often recorded in other carangids, and they usually present a pattern of feeding activity with greater intensity around twilight (Danemann, 1993;Helmer et al., 1995) The circadian rhythms in fish may be associated with various factors such as light, temperature, salinity, and availability of prey or with association of these factors.The light has been suggested as a strong and universal influence, acting powerfully in the synchronization of these physiological and behavioral rhythms (McFarland, 1986;Boujard & Leatherland, 1992;Marques & Menna-Barreto, 2003).Diurnal and nocturnal feeding activity must be related to the difference in the intensity of light, which brings alternation in the period of time of the light and dark phases of the day (Boujard & Leatherland, 1992).Hunter (1968) suggested, however, that with adequate light, as when there is a full moon, some fish are able to feed at the surface at night.In the present study, data were collected during periods of the waxing or waning of the moon, therefore without interference of moonlight in the feeding activity of the

%BM
fish.Fluctuations in temperature and salinity during the day were not reflected in the daily cycle.It may be conjectured, however, that the light, during the photoperiod must act as an abiotic factor responsible for the synchronization between the environmental events and the feeding activity of T. lathami.
The vertical migration shown by the species, which migrates to the surface layer during the night and remains at the bottom during the morning (Saccardo, 1987), may also be synchronized by the cycle of availability of prey.Beyond this fact, the migration of prey may also occur in such a way as to synchronize it to the light/dark cycle (Neilson & Perry, 1990).The feeding phase of T. lathami is the same of those that have been registered in the main catches of this species, denoting the higher vulnerability on these times (Almeida, 2001).Overall, the diet of T. lathami was similar in the two phases of feeding activity, being mainly composed of calanoid copepods, which are the most abundant copepod in the study area (Veja-Pérez, 1993).In relation to Eucalanidae copepods, they are abundant along the year (Hernandez, 1999;Oliveira, 1999), which might explain their occurrence over all the 24hours cycle, both in winter and in summer.
Difference as the occurrence of decapod larvae only at the twilight and nocturnal phases in the winter, for instance, may be related to the phase in the development of these organisms.In the case of the ingestion of cumaceans mainly in summer, it must be associated with their greater abundance in this season, related to the penetration of the cold water of the SACW (Santos, 1992).In the summer the penetration of the SACW leads to the nutrient enrichment of the euphotic zone water, causing an increase in the production of phytoplankton.This primary production sustains a larger biomass of herbivorous zooplankton and offers better conditions for the survival of fish larvae and benthic animals (Pires-Vanin & Matsuura, 1993).Cumaceans present the same pattern of daily vertical migration (Santos & Pires-Vanin, 1999) as T. lathami (Saccardo, 1987), moving upward during the night and descending during the morning.This co-existence explains the prey-predator relationship between cumaceans and T. lathami.
Trophic interactions are complex relationships that involve food, fish, predators, biological process and displacement in water (Manteifel et al., 1978).The vertical migration of many species of crustaceans is due to endogenous rhythm synchronized by light/dark cycle (Neilson & Perry, 1990) that produces the synchronicity between the movement of the prey and predators.
We may point out that there are clear evidences of temporal organization in the feeding activity of T. lathami in the daily cycle.Based on the evening and nocturnal activity, we can suggest the hypothesis of a circadian rhythm for this species, synchronized by the light/dark cycle and/or by the cycle of availability of food.However, to know if this rhythm is endogenous or exogenous, free-running laboratory experiments are needed in order to study the origin of the rhythm.If this feeding rhythm would be detected in constant condi-tions, without any environmental cycle, it could be considered an endogenous rhythm, characterized by an anticipation of cyclic changes, if not, exogenous rhythm is related with a direct answer to an environmental cycle.
The results related with feeding activity pattern of this study can improve the fisheries resolutions for management of this increasing economical resource.

Fig. 1 .
Fig. 1.Sampling area on the southeastern Brazilian shelf and the trawl stations location.

Table 1 .
Sampling information for Trachurus lathami from Brazilian shelf off Ubatuba.
Fluctuation of water temperature at different times of the day, in winter (23 o 35.6´S/ 44 o 59´W) and summer (23 o 51.3´S/ 44 o 47.6´W) 1988, on the shelf off Ubatuba.

Table 2 .
Frequency of occurency (O%), percent mass (M%), and feeding index (FI%) of food items of Trachurus lathami collected in winter and summer from Brazilian shelf off Ubatuba.

Table 3 .
Feeding index of food items of Trachurus lathami collected in winter and summer from Brazilian shelf off Ubatuba.*and **Percent similarity index (PS) > 60%.

Table 4 .
Diet composition of Trachurus lathami expressed by feeding index (FI%) according of lenght classes (mm) of fish collected in winter and summer on Brazilian shelf off Ubatuba.* and ** Percent similarity index (PS) > 60%.Mean of stomach index of fullness expressed as BM% (mass of the stomach content x mass of the fish -1 x 100) of adult (> 115mm) and juvenile (< 115mm) Trachurus lathami by sampling period from Ubatuba shelf, in winter and summer.