the zooplankton, which will raise the survival chances for larval and juvenile individuals. Estuaries are the principal feeding and growing area for early stages of several marine species and few species displace themselves toward estuaries in order to spawn. Conversely, many species spawn on the continental shelf (Chaves & Bouchereau 2000). During processes of habitat shifts, an aggregation phase of many species may occur near estuary mouths, particularly in beach environments, which could be considered an additional recruitment area essential for their life cycle.
This hypothesis can be supported by the occasional occurrence of many engraulids such as Anchoa lyolepis, tungao A. tricolor (Spix & Agassiz 1829) and the clupeid S. brasiliensis 1. The recruitment of these species appears to occur during late summer or early autumn (Godefroid 1996), when many shoals approach shallow waters and leave soon after. This behaviour, according to McFarland (1963), may indicate a reproductive pattern or a simple population aggregation, the purpose of which is still unknown. Mean SL of S. brasiliensis 1 increased continuously from February (40 mm) to May (80 mm) and the same occurred in relation to A. lyolepis, although in a smaller proportion (42–48 mm). As reproduction data for these two species were absent, and capture were restricted to few months, it could be assumed that the studied area is inserted in a migrantory route, acting as a temporary refuge or an orientation site, as concluded by Modde (1980).
Harengula clupeola and T. carolinus seem to recruit all year round, as previously described by Godefroid (1996)., but M. littoralis also presented this pattern, since they are numerous and their presence is constant in all seasons. H. clupeola showed a clear growing pattern in the studied area. Mean SL value increased from 50 to 80 mm until January and a new body-size class entry was observed in the subsequent month. January was the only month when mature individuals were captured (C stage), although juveniles (A stage) were observed only in February. This coincides exactly with the cohort entrance and is an indication of the recruitment. Modde (1980) found for congener H. jaguana two annual peaks of recruitment, one in early spring and another in late summer, with few individuals exceeding 50 mm. He attributed this lack of larger individuals to both reduced catchability and movement outward from the beach. In the present study, individuals larger than 80 mm were not caught perhaps due to the large mesh-size of the seine net or to their absence of the study area. The most acceptable hypothesis is the latter, since
C. FÉLIX. ET AL.
clupeola is a small body-sized species with
maximum length of 170 mm (Figueiredo & Menezes 1978) and not capturing this species would be improbable, since much bigger individuals such as the common halfbeak Hyporhamphus unifasciatus (Ranzani 1842) (133.4 mm) or timucu Strongylura timucu (Walbaum 1792) (290.07 mm) have been caught.
Additionally, T. carolinus and M. littoralis did not present any indication of mean size increment and they were represented only by immature individuals, which could be an indication of continuous spawning activity. This resulted in simultaneously entrance and consequent mixture of many cohorts in beach environments (Mode 1980) that also prevented the visualization of defined cohort patterns. The dominance of immature individuals leads to the assumption that either bigger fish were not caught by the sampling gear or they left the area after reaching some particular size. Fields (1962) and Modde (1980) proposed that T. carolinus of sizes of 60-70 mm and 80 mm, respectively, move offshore to deeper waters. Also Bellinger & Avault (1970) verified that fish up to 100 mm SL class remain in the surf zone. In the present study, mean value of 37.71 mm SL were found for T. carolinus, and the biggest fish captured was 106 mm SL. However, Godefroid (1996), working with a 18 m seine net with smaller mesh- size, caught fish measuring 70-80 mm and only one specimen 90-100 mm size class. Perhaps the same pattern found by Bellinger & Avault (1970) could be applied to beaches at Pontal do Sul, but more samples with different fishing gears are necessary to reach a conclusion about this issue.
Growth patterns could not be identified for the congener T. goodei, which was caught only as juveniles. This result can be explained by the same hypothesis presented for T. carolinus and M. littoralis, however, further studies also need to be conducted to clarify this matter. O. bonariensis is a cold water species that usually migrate from Southern Brazil to warmer waters during winter. High occurrence of this species took place in June and July, appearing sporadically in the other months. Even with few individuals captured during the rest of the year, an increasing size trend could be observed towards summer months, but it could not be considered conclusive because of the low incidence of individuals. Similarly, P. saltatrix, a piscivorous fish, seems to utilize the area for feeding and growth, where it doubled its size (60 to 120 mm). P. saltatrix probably took advantage of the abundant food resource provided by many shoals of clupeiforms.