Weight-length and Length-length relationships for 37 demersal fish species from the Marapanim River , northeastern coast of Pará State , Brazil

Biota Amazônia ISSN 2179-5746 ABSTRACT: Length-length and weight-length relationships are presented for 37 demersal fish species caught in the main channel b of the Marapanim River. The mean allometric coefficient (b) in the weight-length relationship (W = aL ) was 3.04 (±0.32). Positive allometry was predominant (15 species, 42.1%), followed by isometry (13 species, 34.2%), and negative allometry (9 species, 23.7%). The present study represents the first reference of length-length relationships for the northern coast of Brazil.


Introduction
Weight-length relationship is an important tool used in fish biology and stock assessment studies (ABDURAHIMAN et al., 2004).Such relationship allows the estimation of the fish weight using a particular length and may be applied to studies on gonadal development, feeding rate and maturity condition (LÊ CREN, 1951).Nevertheless, parameters in this relationship may vary temporally and/or spatially for a particular species, and require a regular update and estimation for each population separately (ISMEN et al., 2007).Intra-specific variations of the weight-length relationships may be substantial, depending on the period, the population, or the annual differences in environmental conditions (FROESE, 2006).In this context, the present study provides the length-length and weight-length relationships for 37 species of fishes from the subtidal zone of the main channel of the Marapanim River estuary, northern coast of Pará State.

Material and methods
Fish were collected in the estuary of Marapanim River, northeastern region of Pará State.This estuary is part of the Salgado region of Pará State, near the mouth of Pará River, approximately 160 km from the mouth of the Amazon River (Figure 1).Monthly samplings between August 2006 and July 2007 were carried out using a bottom trawl net with doors (model "Wing Trawl").Tows occurred during daytime ebb tides at depths between 1.5 and 3.0 m.Specimens were identified to species level based on the pertinent literature.Measurements of total length (0.1 cm), standard length (0.1 cm) and total weight (0.01 g) were recorded.
Length-length relationship was calculated using the equation TL = a + bSL, where TL is the total length (cm) and SL the standard length (cm).The weight-length b relationship was calculated using the equation W = aL (PAULY, 1984) with the data transformed into log W = log a + b log L, where W is the fish weight (g) and L is the total length (cm).To test whether the value of b in the weight-length relationship was significantly different from the value of isometry (b = 3), the Student t-test was applied with a confidence level of ± 95% (α = 0.05), according with Sokal and Rohlf (1987).

Results
A total of 19.886 specimens covering 17 families and 37 species was analyzed.Information regarding the species identification (Family and Species), length data, weight data and parameters of the relationships are shown in Daniel Ferraz e Tommaso Giarrizzo

Discussion
According to Froese (2006), positive allometry (b > 3) suggests that large specimens have increased in height or width rather than in length, isometry (b = 3) indicates that small specimens in the sample have the same shape and condition of large specimens; negative allometry (b < 3) indicates that large specimens have changed the shape of the body to become more elongated specimens.Estuaries have high primary productivity, shelter and refuge areas against predators, natural nurseries for many species of fish and large availability of food (LAEGDSGAARD & JOHNSON, 1995, 2001;BARLETTA-BERGAN et al., 2002a, b;BARLETTA et al., 2003).Thus, it is expected that most of the species analyzed show isometry or positive allometry in their respective type of growth, which is due to the vast availability of food provided by the habitat.
The linear relationship resulting from the weight-length relationship between Log (a +1) vs b can be used to identify possible outliers in the sample (FROESE, 2000;STERGIOU & MOUTOPOULOUS, 2001) (Figure 2).Most species (91.9%) showed b values distributed between 2.5 and 3.5.According to Ricker (1975) In a study carried out in the intertidal creeks of Curuçá Estuary, northeastern coast of Pará State, approximately 20 km from the Marapanim Estuary, Giarrizzo et al. (2006) analyzed 40 species of fishes.Our estimates of b were lower than those of Giarrizzo et al. (2006) for 10 species and higher for eight species (Table III).Most of the species analyzed by Giarrizzo et al. (2006) showed a positive allometric growth (21 species), which corroborate with the present study.This result may be associated with the typical ichthyofauna of estuarine habitats, with high numbers of juvenile or small adults that inhabit environments with high food availability.Joyeux et al. (2008) in a study conducted along the Brazilian coast have found b values to be higher than in the present study for most of the species (Figure 3; Table III).Only species caught in the estuary of Curuçá River were compared with the present study and those from other regions were not considered.The latter study showed identical sampling methodology with the present study, and used trawl nets of the same length and mesh-size.Such procedure allowed a more efficient comparison by removing the effect of the variability of the estimates of b, which could be attributed to differences in the gear and the selectivity of the net.Of the 28 species in both studies, for only six species our estimates of b were higher than those found by Joyeux et al. (2008).The observed differences can be explained by a number of factors including food availability, number of specimens and variations in the length range of the populations sampled (PAULY, 1984;WEATHERLEY & GILL, 1987).For Giarrizzo and Krumme (2006), the morphology of the estuary mouth and the proximity to the plume of the Amazon River are important factors that influence the estuarine ichthyofauna in northern Brazil.The estuary of Marapanim River is located approximately 160 km away from the mouth of the Amazon River, while the estuary of Curuçá River is located approximately 20 km closer.In addition to the larger area of mangrove cover, the proximity between the Curuçá River estuary and the mouth of the Amazon River may provide higher nutrient input to the estuary, allowing greater availability of food for the fishes compared with the estuary of Marapanim River.According to Tesch (1971), biological factors typical of the species are also relevant, such as growth phase, degree of stomach fullness, gonad maturity, sex, length frequency, health, and conservation techniques.In this regard, any comparison involving lengthlength or weight-length relationships should be analyzed with caution, as many factors may contribute to the great variability observed.
The results shown in this study represent the first reference of length-length relationship for the northern coast of Brazil and may contribute to a better assessment of fish stocks in estuarine areas.
, values of b outside this range are considered atypical.Only B. vaillantii (b = 1.78) was below this range, and C. faber (b = 3.66) and Mugil rubrioculus (b = 3.51) were both above this range (Figure 2) in the estuary of Marapanim River.

Figure 3 .
Figure 3. b (JOYEUX et al., 2008) vs b (this study) of the 28 species present in both studies for comparison between the two areas (Curuçá River Estuary and Marapanim River Estuary) of the northeastern coast of Pará State.

Table II .
Summary of weight-length relationship describing length characteristics (cm), weight characteristics (g) and the parameters of weight-length relationship.TL min: minimum total length; TL max: maximum total length, SE: standard error; TW min: minimum total weight, TW max: maximum total weight; Biota Amazônia

Table III .
Table of comparison between the present study and other studies carried out on the northeastern coast of the state of Pará, Salgado micro-region of Pará State.n: number of individuals analyzed; TL min: minimum total length; TL max: maximum total length; b: allometric coefficient of the regression, SE: Biota Amazônia