1 INTRODUCTION

The mathematical relationship between length and weight of fishes is a practical index suitable for understanding the survival, growth, maturity, reproduction and general well-being, and is often used to characterize life history and make morphological comparisons between different fish species or populations (Jennings et al., 2001; Xue et al., 2011). Since the last few years the landing records of marine catfishes has steadily declined in Pakistan, based on regular monitoring of local exporters and suppliers (authors' personal communication). The length-weight relationships (LWRs) of marine catfish species reported by various authors from different regions are summarized in Table 1. In this study, LWRs have been established for eleven marine catfish species captured from the northern Arabian Sea coast Pakistan.

2 MATERIAL AND METHOD

Specimens were sampled by commercial vessels from December 2014 to November 2015, using bottom trawls with various mesh sizes. The data from 995 specimens were used to calculate LWRs. Measurement of large specimens was made on the spot and for further biological studies specimens were transported in ice boxes to the fisheries laboratory at the Centre for Excellence in Marine Biology (CEMB). All the species were identified using FAO species catalogues (Fischer and Bianchi, 1984; Kailola, 1999) or FishBase. Each individual was measured for total, fork and standard lengths in centimeters (cm) and weighed in grams. Pooled data were used to calculate LWRs, using the parabolic equation W=aL^b, where L is length (cm), W is the weight (g) of fish. The intercept 'a' and growth coefficient 'b' were analyzed after transforming data on natural log (ln) equation: W=L_na+bL_n and calculating 95% confidence limits (Le Cren, 1951; Froese, 2006).

3 RESULT

Data from 995 specimens were used to calculate length-weight relationships. The regression parameters (a, b, R² and confidence limits) are summarized in Table 2. It is interpretable from the R² value 0.8 to 0.9 that there is strong linearity between the parameters. The values of slope b indicated that five species' growth was positive allometric (b > 3) and rest of the six species' growth was negative allometric (b < 3). The length-weight relationship of four species is reported for the first time and maximum length for three species should also be added to FishBase.org.

4 DISCUSSION

In this study, calculated b values were within the proposed range of 2.5–3.5 (Froese, 2006). The b values of A. gagora, N. caelatus, P. tenuispinis, O. militaris, P. lineatus were > 3, and indicated positive allometry. In N. bilineata, P. dussumieri, S. sona, N. thalassina, B. mino and P. limbatus, on the other hand, values were < 3, indicating negative allometry. The b values above 3 are the result of large specimens that increase in girth or width (Froese, 2006). The b-values for P. dussumieri and N. caelatus were close to those in the same species reported from West Bengal (Dutta and Hazra, 2013). The parameter b varied significantly among the species and this variation within a similar habitat may be ascribed to differences in the physiology of the fishes (Le Cren, 1951).

It is concluded that, this study has contributed to the knowledge of catfish species from the northern Arabian Sea coast of Pakistan and that the data should be utilized for better management and conservation of these important fisheries.

5 ACKNOWLEDGEMENT

The assistance of the Director, Center of Excellence in Marine Biology, University of Karachi is greatly acknowledged.