Indian Journal of Geo Marine Sciences Vol. 47 (01), January 2018, pp. 135-140
Biometric studies on spotted sardinella Amblygaster sirm (Walbaum, 1792) (Pisces: Clupeidae) occurring along Andaman coast, India
1Sukham Monalisha Devi, *1Ashok Kumar Jaiswar, 1Rajan Kumar, 2Mohammad Irfan Ali, 2Sajin Velakkandy, 3Swapnil Shirke, 1Shrinivas Jahageerdar & 1Sushanta Kumar Chakraborty
1Fisheries Resource Harvest and Post-Harvest Management Division, Central institute of fisheries education, Mumbai-61, India
2Central Island Agricultural Research Institute, Port Blair-01, Andaman and Nicobar Islands, India
3Fishery Survey of India, Port Blair-01, Andaman and Nicobar Islands, India
*[E-mail: akjaiswar@cife.edu.in]
Received 29 September 2015 ; revised 30 October 2016
All the morphometric variables of Amblygaster sirm show significant correlation between them except head length with showed correlation coefficient of less than 0.80 with all other morphometric variables. Liner regression between standard length and other morphometric variables showed R2 value of > 0.9 except for HL and BD were the values dropped below 0.75. Pairwise linear regression of log transformed morphometric variables with log-standard length shows negative allometric growth for HL and positive allometric growth for body depth with standard length. Descriptive statistics for morphometric and meristic characters were found to be in agreement with previous work. Spearman rank correlation shows insignificant relation between meristic characters and standard length. L-W relationship for males and females were established as W=0.000007L3.054 and W=0.000001L3.370 respectively. Relative condition factor (Kn) for females (mean=) were found to be higher than males (mean=) throughout the year with no major seasonal fluctuation.
[Keywords: Amblygaster sirm, biometric, L-W relation, relative condition factor]
Introduction
India is one of the major fishing countries in the world with 8,129 km coast line and 2.02 km2 EEZ1. Clupeoids fishes contribute nearly 50 % of the pelagic fish landings. They are most important marine fishes across the globe, both in terms of the quantum of its production and its socioeconomic importance to coastal community. Spotted sardinella, Amblygaster sirm is characterized by slender body and a series of 10-20 gold spots down the flank. They are distributed all along Indo-West pacific. It is a coastal and pelagic schooling fish2. Rajan et al. reported 19 clupeid species from Andaman waters3. A. sirm is the most important clupeid species from Andamans and landed throughout the year. Major landing is from ring seines targeting the shoals. A minor proportion is also coming from gill nets. A. sirm is widely used as bait in tuna long lining apart from being consumed fresh
by locals. Previous estimates of morphometric and meristic characters of A. sirm were done by Chan4, Losse5, Whitehead2 and Vijayakumaran et al.6. Length-weight relationships have been earlier worked out by Athukoorala et al.7 from Sri Lankan waters and Abdussamad et al8. from Gulf of Mannar ecosystem.
Studies on similar lines have not been carried out for A. sirm from Andaman waters. Current investigation is an attempt to quantify morphometric and meristic characters of the species, their inter-relationship, growth pattern, L-W relationship and condition factor of the given species.
Materials and Methods
The present study was based on samples collected from Junglighat and Dignabad Fish Landing Centre of Port Blair, Andaman and Nicobar Islands. Apart from that, samples were also collected form fish markets. A
total of only 40 specimens of A. sirm were examined for morphometric and meristic studies as specimens with narrow band of size range were only available.
Eight morphometric characters namely standard length (SL), fork length (FL), total length (TL), head length (HL), pre-dorsal length (PDL), pre-anal length (PAL), pre-ventral length (PVL) and body depth (BD) were recorded for the current investigation. The characters are selected considering their ease in recording even at landing centre. Seven of the morphometric variables were scaled to standard length and a descriptive statistics including mean, maximum, minimum, standard deviation and coefficient were tabulated for comparative study with past literature. Similarly, descriptive statistics for eight meristic characters namely dorsal fin rays (DF), ventral fin rays (VF), Anal fin rays (AF) , pectoral fin rays (PF), caudal fin rays (CF), belly scutes (SC) , gillrakers (GR) and branchiostegal rays (BR) were also tabulated which includes median, mode, frequency of mode, minimum, maximum and coefficient of variation. A Pearson correlation matrix for morphometric variables was generated to see nature of relationship between them. For meristic characters a Spearman rank correlation was calculated with standard length to explore the relationship between length of fish and meristic counts. Linear relationship of different morphometric variables with standard length was established using least square method described by Laevastu9 and Snedecor and Cochran10. Pairwise linear regression of log- transformed morphometric variables with log- standard length were carried out and regression coefficient were assessed for allometric growth of the considered variable with respect to standard length11,
12. Length weight relationships were established for males, females and pooled data separately using the formula of LeCren13. For estimating L-W relationships samples were collected over a period of one year covering 588 male specimens and 616 female specimens. Analysis of covariance (ANCOVA) was carried out to assess the difference in mean log-weight of females adjusted for covariate (log-length) and to test the homogeneity of regression slopes between males and females at 1% and 5% level of significance. A student t- test was carried out for calculated “b” value against the value of 3 to test the growth pattern of males and females against the null hypothesis of isometric growth (H0: b=3) both at 1%
and 5% level of significance. A month wise relative condition factor13 for males, females and pooled data
were estimated to assess the differences across sexes and months of the year. All the statistical analysis was performed using MS- Excel, PAST 3.0 and statistical software package Statistica.
Results
Descriptive statistics of seven morphometric ratios reveals maximum coefficient of variation for BD/SL (4.825%) followed by HL/SL (4.321) whereas minimum were recorded for FL/SL (1.216) and TL/SL (1.369) (Table: 1). A correlation matrix (Pearson) of eight morphometric variables showed significant correlation between the variables. A high correlation of > 0.94 was observed between SL, FL and TL. Head length was found to be least correlated with other morphometric variables with highest recorded coefficient of 0.754 with PDL and lowest of 0.594 with FL. Body depth also showed relatively lower correlation (<0.9) with the size variables (SL, FL and SL) (Table: 2).
Performed linear regression between SL (X- variable) and other morphometric variables (Y- variables) shows a near linear fit with highest R2 value recorded for FL (0.9272). A lower R2 value of <
0.75 was estimated for HL (0.4057) and BD (0.7247) (Fig. 1). A pair wise linear regression between SL (X- variable) and other morphometric variables (Y- variables) after log transformation showed a „b‟ value of near unity for TL, FL, PAL, PVL and PDL.
Regression coefficient (b) values of near unity show isometric growth of the variable with respect to standard length. A substantially lower value of „b‟
(0.745) has been recorded in case of head length (HL) showing negative allometric growth. On the other hand body depth (BD) recorded a positive allometric growth with respect to standard length with a
regression coefficient of 1.493 (Fig. 2).
Descriptive statistics of eight meristic characters recorded highest variation in case of gillrakers followed by anal fin rays with respective coefficient of variation of 7.33% and 6.99%. Ventral fin rays (8) and branchiostegal rays (5) remained constant throughout the examined specimens (Table: 3).
Table 3- Descriptive statistics for meristic counts of Amblygaster sirm Meristics N Median Mode
Freq. of
mode Min. Max. Losse5 Chan4 Vijayakumaran et al. 6
DF 40 18 18 35 18 19 17-19 17-19 16-18
PF 40 17 17 31 16 17 16-17 16-18
VF 40 8 8 40 8 8
AF 40 17 17 22 15 21 20-22 16-18 17-20
CF 40 25 24 20 24 26
SC 40 31 31 15 29 34 29-32 29-31
GR 40 36 Multi 6 33 41 31-42 38-42 38-42
BR 40 5 5 40 5 5
Spearman rank correlation coefficient (rho) between meristic variables and standard length (size variable) were found to be insignificant both at 1%
and 5% level of significance level. None of the variable recorded a rank correlation coefficient of >
0.30 (Table: 4).
Length-weight relationship and condition factor Length-weight relationship for males, females and pooled data were established as:
Male: W=0.000007L3.054 Female: W=0.000001L3.370 Pooled: W=0.000003L3.216
Same power function could be written in linear for after log-transformation as:
Male: LogW=-5.125 + 3.054 LogL Female: LogW=-5.872 + 3.371 LogL Pooled: LogW=-5.507 + 3.216 LogL
Fig.1. Relationship (Linear fit) between morphometric variables and standard length
Fig.2. Relationship (Linear fit) between log- transformed morphometric variables and log standard length
Table 4- Spearman rank correlation coefficient of meristic counts with standard length for A. sirm
Meristics Corr. Coeff. With SL Sig. (2-tailed)
DF -0.036 0.825*
PF -0.156 0.337*
VF NA NA
AF -0.024 0.883*
CF 0.213 0.187*
SC 0.024 0.883*
GR -0.038 0.815*
BR NA NA
*Insignificant at 1% and 5% of level of significance
Power relationship of length and weight for males (Fig. 3), female (Fig. 4) and pooled (Fig. 5) is being presented in form of scatterplots.
Analysis of covariance (ANCOVA) for log transformed length and weight indicated no difference in mean weight of males and females (adjusted for covariate length) but same is not true for regression slope. Regression slope for males and females are significantly different at both 1% and 5% level of significance as indicated by p (0.0077)<0.01 (Table:
5). The regression slope between Log W and Log L for both male and female is depicted in a scattergram (Fig. 6). A t-test conducted separately for male and female to explore the growth pattern of the same.
Female showed a positive allometric growth both at 1
% and 5% level of significance indicated by higher calculated t-statistic (5.029) than table value. On the other hand male showed isometric growth as calculated t-statistic (0.573) was smaller than table value both at 1% and 5% level of significance.
Table 5- ANCOVA table for test equality of means (Log w), adjusted for covariate and for homogeneity of slopes
Sum of sqrs. df
Mean square F
p (same) Adj. mean: 0.0018 1 0.0018 0.845 0.358 Adj. error: 2.5241 1201 0.0021
Adj. total: 2.5258 1202 Test for homogeneity (equality) of slopes
F : 7.12
p (same) 0.0077 (<0.01 shows inequality of slopes) Fig.4. Length-weight relationship for A. sirm (female).
Fig.3. Length-weight relationship for A. sirm (male).
Fig.5. Length-weight relationship for A. sirm (pooled).
Relative condition factor (Kn) for male, female and pooled data were estimated in current study. For males the Kn ranges from 0.356 to 1.70, whereas for females the value was estimated between 0.936 and 2.163. The pooled data showed the range of 0.350- 1.683. The mean relative condition factor for female (1.349) is higher than that of male (1.077). A month wise exploration of Kn indicates no major fluctuation across the year for both males and females (Table: 6).
Table 6- Month wise relative condition factor (female, male and pooled) A. sirm
Mean Relative condition factor
Pooled Female Male
Month N Kn N Kn N Kn
Jan 105 1.0632 50 1.3619 55 1.1021 Feb 103 1.0939 32 1.4402 71 1.1174 Mar 116 1.0618 42 1.3745 74 1.094 Apr 109 1.0362 65 1.3394 44 1.0653 May 104 1.055 62 1.3648 42 1.0858 Jun 108 1.0098 76 1.2986 32 1.0738 Jul 111 1.0234 63 1.3227 48 1.0558 Aug 94 1.0274 65 1.3574 29 1.01 Sep 68 1.0249 40 1.3153 28 1.0689 Oct 73 1.0554 36 1.3796 37 1.075 Nov 104 1.0326 44 1.3386 60 1.06 Dec 109 1.0488 41 1.3722 68 1.0707 Total 1204 1.0447 616 1.3495 588 1.0779
Min. 0.3508 0.9362 0.3569
Max. 1.6826 2.1635 1.7316
Discussion
All the considered morphometric ratios showed relatively lower coefficient of variation (Table: 1) which could possibly be due to narrow band of size range (TL=19.1-24.0 cm) explored during the present study. The narrow size range (constraint of present
study) of the sample specimens were owing to the fact that A. sirm fishery in Andamans is of ring seine type targeting shoal of only a specific size range.
Nevertheless, the mean value of the morphometric ratios presented in current work (Table: 1) were in broad agreement with previous works4, 5. Most of the morphometric variables showed linear relationship with standard length with high R2 value except head length and body depth (R2 <0.75). Pairwise regression of these two variables after log transformation gave a regression coefficient quite different from unity, showing strong allometric growth with respect to standard length explaining poor R2 value for linear regression model and their lower correlation coefficient with longitudinal size variables (SL, FL and TL). Head length in current study showed negative allometric growth which is common in several group of fishes. The rate of growth of head length slows down as fish grows bigger in length with age14. Present study also recorded a positive allometric growth in body depth. A. sirm, especially females tend to become broader as they grow in length/age which could be easily visualized by seeing male and females of same length simultaneously. The increasing broadness is possibly more evident once the fishes surpass its length at first maturity (Lm).
Several other fishes were also found to show positive allometric growth for morphometric variables explaining body width15. The descriptive statistics for eight considered meristic characters were found to be in agreement of with previous literatures4, 5, 6. Poor spearman rank correlation coefficient (<0.30) for all the meristic characters (Table: 4) with SL shows no definitive increasing or decreasing pattern of meristic counts with increase in length. This implies that the size of the fish do not dictates the carried meristic counts.
Length weight relationship showed differential growth pattern for males and females. An isometric relationship was observed for males whereas females grow in allometric fashion. Athukoorala et al.7 reported positive allometric growth for both males and females from waters of Sri Lanka. Abdussamad et al.8 estimated „b‟ as 3.207 (for pooled data) for A.
sirm from Gulf of Mannar which is quite similar to the findings of current study (3.216). Positive allometric growth for females appears quite obvious owing to broader body than their male counterparts of same length. Here, it is worth mentioning that majority of the specimens considered for current estimation were of size possibly greater than reported length at first maturity (Lm) of 16 cmby Abdussamad et al.8 and hence strong positive allometry seems quite
Fig.6. Regression slope of length-weight relationship for males (red) and females (black) of A. sirm
justifiable. Relative condition factor of a fish is affected by several factors like feeding intensity and gonad development13. During current investigation relative condition factor (Kn) for females were higher than that of males for the entire year. Similar condition was also reported for other clupeids16. Relative condition factor for both males and females showed no prominent difference across different months of year. Such a condition could be attributed to the asynchronous breeding behavior of the species, spanning over most parts of the year8. To add to this, similar feeding intensity throughout the year (observed during current study) could possibly be accountable for similar Kn for all the months.
Conclusion
Current study reveals that most of the morphometric variable follow isometric growth pattern barring head length and body depth. They show negative and positive allometric growth with respect to standard length respectively. A sex segregated morphometric assessment along with larger sample size with bigger size range could fine tune the current estimates in future. Females show positive allometric growth different from isometric growth pattern of males. Condition factor of females were higher than males. Kn for both males and females do not show any seasonality. A detailed biological estimate on feeding and reproductive aspect is required to unravel the underlying reason for such pattern.
Acknowledgements
The authors are grateful to the Director, ICAR- CIFE, Mumbai; Director, CIARI Port Blair and Scientist In-charge, Fisheries Division, CIARI, Port Blair for providing funds and facilities to conduct this study.
References
1. Pillai, N.G.K., Pelagic Fisheries. Handbook of Fisheries and Aquaculture, (ICAR, New Delhi) 2011, pp. 66-69.
2. Whitehead, P.J.P., FAO species catalogue, Clupeoid Fishes of the world Vol.7, (FAO, Rome) 1985 pp. 303.
3. Rajan, P.T., Sreeraj, C.R., and Immanuel, T., Fishes of Andaman and Nicobar Islands: A checklist, Jour.
Andaman Sci. Assoc., 18(1) (2013) 47-87.
4. Chan, W. L. Y., A systematic revision of the Indo- Pacific clupeid fishes of the genus Sardinella (Family Clupeidae), Jpn. J. Ichthyol., 13 (1965) 1-39.
5. Losse, G. F., The elopoid and clupeoid fishes of East African coastal waters, Jour. East African Nat. Hist.
Soc., 27(2) (1968) 77-11.
6. Vijayakumaran, K., Diwakar, K., and Achayya, P., Occurrence of spotted Sardinella along north Andhra Pradesh coast. Mar. Fish. Info. Serv., 139 (1995) p.10.
7. Athukoorala, A. A. S. H., Bandaranayaka, K. H. K., and Haputhantri, S. S. K., A study on some aspects of reproductive biology and population characteristics of Amblygaster sirm in the west coast of Sri Lanka, Int.
J. fish. aquat. stud., 2(4S) (2015) 41-45
8. Abdussamad, E. M., Pillai, N. G. K., Habeeb Mohammed, O. M. M. J., and Jayabalan, K., Sardines of the Gulf of Mannar ecosystem-fishery and resource characteristics of major species, Indian J. Fish., 57(4) (2010) 7-11.
9. Laevastu, T., Manual of methods in fisheries biology:
Research on fish stocks. FAO manuals in fisheries science, 4 (1965) 1-51
10. Snedecor, G. W., and Cochran, G. W., Statistical methods (6th ed.), (Oxford and IBH publishing Co. New Delhi) 1967 pp. 593
11. Huxley, J. S., Problems of relative growth, (Metheun and Co., London) 1932 pp. 276
12. Klingenberg, C., Multivariate allometry In: Advances in morphometrics, (Springer US) 1996 pp. 23-49.
13. LeCren, E.D., The length-weight relationship and seasonal cycle in gonad weight and condition in the Perch Perca fluviatilis, J. Anim. Ecol., 20(2) (1951) 201-219
14. Chioma, G. N., Adejumo, A. O. and Olumoh, S., Allometric and isometric growth of external body parts of Auchenoglanis occidentalis (Pisces: bagridae), Sci.
Foc., 12(2) (2007) 76-82.
15. Klingenberg, C. P., and Froese, R., A multivariate comparison of allometric growth patterns, Syst.
Zool., 40(4) (1991) 410-419.
16. Zaki, S., Jayabalan, N., Al-Kiyumi, F., Al-Kharusi, L., and Al-Habsi, S., Maturation and spawning of the Indian oil sardine Sardinella longiceps Val. from the Sohar coast, Sultanate of Oman, J. Mar. Biol.
Assoc., 54(1) (2012) 100-107.
