Feeding ecology, prey selection and growth performances of gold spot mullet, Liza parsia (Hamilton, 1822) in extensive brackish water farming system of Indian Sundarbans

A study was aimed to focus on the feeding ecology with prey selection and growth performances of gold spot mullet (Liza parsia) in extensive polyfarming system. Three ponds (0.7-0.9 ha) were brought under observations which continued for 10 months (JanuaryNovember, 2015). Estuarine water containing wild fry of various species were allowed to enter in three impoundments. Commercial feed or fertilizer was not applied following common practice. Exchange of 20-30% water was carried out during lunar cycles. Water temperature and water depth were average 29.90C and 120 cm. Order of dominance of prey groups in water was chlorophyceae, bacillariophyceae, myxophyceae, copepod and dinoflagelats. Tendency towards higher feeding intensity in terms of stomach fullness was observed as the fish grew. Order of dominance of prey groups in stomachs were myxophyceae, chlorophyceae, bacillariophyceae, dinoflagelatsand copepods while myxophyceae followed by bacillariophyceae was actively selected. Copepods were selected for first two months and negative selection of all zooplankton groups were noticed afterwards. L. parsia attained 58.94±2.02g (15.2±0.19cm) at harvest with slope value of length-weight relationship (b=3.028) and condition factor (K=1.26±0.11).Good growth and health condition in the present study can be attributed to the low production farming system with comparatively lower stocking density providing greater food and space for the growing fish.


Introduction
The gold spot mullet Liza parsia(Hamilton 1822), belonging to the family Mugilidae is a catadromousfish and widely distributed in the coastal waters of tropic, sub-tropic and temperate regions extending from 42º N to 42º S [1].With its good market price and consumer preference, abundant availability of seeds, non-carnivorous food habit, gold spot mullet is easily cultivable in shrimp growing areas and forms commercially important candidate for polyculture with shrimp and other mullets [2][3][4][5].Gold spot mullet, popularly known as 'parsia' in Indian subcontinent, can tolerate wide range of environmental fluctuation [6].
In India, L. parsia occurs in marine, shallow coastal water, shallow coastal lakes and brackish water estuaries and is reared in brackish water tide fed ponds in West Bengal [7,8], in the estuaries and brackish water lakes of Kerala [9,10] and in Lake Pulicat [11,12].In West Bengal, the low-lying lands near estuaries and deltaic areas enclosed by embankments called "bheries" are used for traditional finfish cultivation mostly for mullets, especially during rains [13].In bheries, large numbers of fish and shrimp seeds brought in through tidal water and partial stocking are reared for a period of 6-7 months [14] where gold spot mullet is an important component.Knowledge on species niche and growth potential in such extensive farming ecosystem is necessary for proper management and formulation of strategies for sustainable production improvement.Studies on the food and feeding habit indicate the species niche in ecosystem, their food preference and food spectrum asish177@gmail.comoverlaps.Feeding behavior at the level of prey selection can have implications at the individual [15], population [16] and community levels [17].But information on food spectrum and prey preferences of parsia in particular is scanty.
Growth performance of fish is one of the most important criteria for selection as a candidate species for farming.Available reports regarding growth of parsia is highly variable from farming systems.L. parsia fingerlings (1.75g) were grown to 52.41 and 66.1g in 18 months with Liza tade at ratios of 2:1 and 4:1 at overall stocking density of 25000/ ha in West Bengal coast [7].Ali et al [2] reported growth of L. parsia fry (0.20g) up to 30.45g at stocking density of 10000nos/ha in 120 days of culture with Penaeus monodon at stocking density of 40000nos/ha in single crop system.He also reported that parsia fry were grown upto47.68gat similar initial size and stocking density in 225 days of rearing with P. monodon at stocking density of 20000nos/ha in double crop pattern.Much higher growth was reported by Biswas et al [8] where L. parsia fingerlings (10.68±0.56)attained 54.02±2.11gat stocking density of 2000nos/ha in 180 days polyculture with Mugil cephalus (4500nos/ha), L. tade (1500nos/ ha) and P. monodon (20000nos/ha) in tide fed polyculture system.There is scarcity of information regarding growth of parsia in extensive farming system traditionally practiced in Hooghly-Matla estuarine complex popularly known as Sundarbans.
The present study was aimed to assess feeding strategy along with prey preferences and growth performances of parsia reared in extensive polyculture system of Indian Sundarbans with a view towards optimum utilization of this high value species in improved polyculture utilizing all the potential food of the water bodies without much competition.

Materials and Methods
Gopalnagar Dakshin village coordinating 21.802-21.807°N,88.296-88.298°E of Pathapratima Block, South 24 Parganas, West Bengal, India situated at the bank of Hatania-Doania River was selected.Three ponds (0.7-0.9 ha) were brought under observations which continued for 10 months (January-November, 2015).Initially, the ponds were dewatered and sun dried.Liming was done over the dried pond bottom using Lime Stone Powder (LSP) @500 kg/ha during first week of January.Ponds were filled with high tidal saline water (18.5ppt) up to a depth of 120 cm during second week of January after filtration through traditional bamboo screen.Small fry can enter through traditional bamboo screen during water exchanges but exit of bigger fishes is restricted.As seeds of L. parsia remain available in north-east coast of India during the month of January-March [2,4,18], entry of gold spot mullet fry along with other species was anticipated.Following traditional practice, no fertilizer or supplementary feed was used and fishes grew depending on natural food.Natural food was brought in through exchange of 20-30% water during lunar cycles.Pond water and fish samples were collected monthly mid-February onwards from three ponds and analyzed.
Stomachs of ten fishes from each impoundment every month were removed intact and feeding intensity based on fullness of stomachs was recorded by visual estimation and classified as gorged, full, ¾ full, ½ full, ¼ full, Little and empty [21].Stomach contents were preserved in fixed volume of 5% buffered formalin.Planktonic constituents of fish stomachs were identified and counted following direct census method [18,20] using Sedgwick-Rafter counting cell and classified in same way as pond water.Additionally, organic matter and sand and mud particles were evaluated as major stomach constituents.Numeric percentages of each group were calculated.
Ivlev's Electivity Index [22] was used to determine prey preferences.Percentage compositions of food items in stomach were compared with that of studied pond water using the following formula: Where, r = percentage of dietary item in ingested food, p = percentage of prey in the environment.
Gravimetric data like total length (TL, cm) was recorded with a slide caliper, while body weight (W, g) was measured using a digital electronic balance prior to dissection for stomach analysis.Ten fishes from each three ponds were collected monthly mid-February onwards i.e. 30 fish in a month and total 300 fish were analyzed throughout the study period.The following formula was used to determine daily weight gain (DWG): Where W f and W i are the average final and initial weight in time t.Specific growth rate (SGR) was determined using the conventional equation:

Feeding ecology, prey selection and growth performances of gold spot mullet, Liza parsia(Hamilton, 1822) in extensive brackish water farming system of Indian Sundarbans
Where W f and W i are the average final and initial weight in time t.
Length-weight relationship was determined using the following mathematical relationship proposed by Pauly [23]: Where W is fish weight (g), TL is total length (cm), 'a' is the proportionality constant and 'b' is the isometric exponent.The parameters a and b were estimated by non-linear regression analysis.
Fulton's condition equation was used to find out the condition factor [24]: Where K is the condition factor, w ̅ is the average weight (g) and (TL) is the average total length (cm) Differences in growth parameters of fish among ponds were determined by analysis of variance with the General Linear Model procedure using SPSS for Windows v.17.0 program me (SPSS Inc Chicago IL USA).Duncan's Multiple Range Test [25] was used for comparison of data.All data are expressed as mean ± standard error (S.E.).
Percentage occurrences of planktonic and other suspended food components in pond water are presented in figure 1.The most abundant phytoplankton groups in three ponds according to the order of dominance were chlorophyceae, bacillariophyceae and bymxophyceae.The most abundant genera found under chlorophyceae were Enteromorpha, Ulothrix, Pediastrum, Chlorella

Feeding ecology, prey selection and growth performances of gold spot mullet, Liza parsia (Hamilton, 1822) in extensive brackish water farming system of Indian Sundarbans
Feeding intensity of gold spot mullet in terms of stomach fullness is presented in table2.Overall, only 6.6% fishes were found with empty stomach considering the whole study period.Most of the fish stomachs were observed with ¼ full (21.88%),½ full (18.40%)and ¾ full (17.99%)stomachs indicating moderate feeding.Highest gorged stomachs was recorded during June (22.2%)and lowest during October (1%) whereas highest (14.8%) and lowest (0%) empty stomachs were found during November and August, respectively.A tendency towards higher feeding intensity was observed as the fishes grew as gorged, full and ¾ full stomachs were gradually increased and ½ full, little food and empty stomachs gradually decreased.Percentage occurrences of food materials observed in the gold spot mullet stomachs are presented in figure2 and 3.The dominant phytoplankton groups in the fish stomach according to the order of dominance were myxophyceae (19.37 in November -39.86% in June; 29.33±1.82%),chlorophyceae(13.04inJuly-21.54 in September;17.30±1.05%)and bacillariophyceae (4.45in April-13.61%inNovember; 9.02±1.05%).Most abundant zooplankton groups were dinoflagelats (1.14 in September -6.37% in October; 2.87±0.55%)and copepod (0.12 in April -4.89% in February; 2.23±0.62%).Among non-planktonic contents, higher plant materials (9.85 in August-19.03% in November; 13.38±1.01),fish and shrimp parts (0.12% in April-3.19% in May; 1.00±0.29%),debris-sand-mud (19.00inFebruary-32.00% in August; 24.87±1.34%)were also present.Month wise electivity index (E)of different food categories are presented in figure 4.Gold spot mullet showed highest preference towards myxophyceae with significant positive E value of-0.37±0.08 ranging between -0.25 and-0.53duringOctober and May, respectively.Bacillariophyceae was found to be second most preferred group with significant positive E of -34±0.06 which varied between-0.22during May and -0.49 during September.E for chlorophyceae ranged between -0.02 to -0.40 during November and July, respectively indicating third most preferred (E=-0.12±0.08)prey group.Copepods showed positive value of E during early months which gradually decreased as the fish grew and varied between-0.17to -0.92 and mean value (-0.46±0.30)indicated negative selection.E for fish and shrimp parts and dinoflagelats ranged between -0.23 to-0.97 (-0.68±0.16)and

Plackett-Burman Design for the Screening of Media Component for Anthracene Degradation by Sphingobium Yanoikuyae Strain ANT3D
Copyright: © 2016 Rajpara, et al.

Discussion
Water quality parameters in the studied extensive ponds were within the optimum ranges for brackish water aquaculture [2, 26 and27].Concentrations of toxic metabolites like nitritenitrogen (NO2-N)and ammonia-nitrogen (NH4-N) remained lower than the critical level and concentrations of nutrients like nitrate-nitrogen (NO3-N) and phosphate-phosphorous (PO4-P) was much lower than fertilized ponds reported from Sundarbans [4,28] and corroborated with those reported earlier by Mondal et al [29] from the same location.Lower nutrient concentrations in the studied extensive system may be attributed to complete dependence on natural productivity without any additional input.
Order of dominance of the planktonic groups in the ambient water in the present study corroborated with that reported from the Hooghly estuary and shrimp farming ponds in Sundarbans [28].Extensive farming can be considered as representative of the natural environment as it depends only on the natural productivity.Existence of planktonic community structure resembling the natural environment in such farming system can be expected.
Gold spot mullet in tide fed extensive farming systems coexists with other herbivorous and carnivorous fishes.Carnivores like Lates calcarifer, Therapon jarbua, Eleutheronema tetradactylum, Glossogobius giuris and many more gains entry in extensive system during tidal water exchange [8].As smaller fishes are more vulnerable, they would rather feed more cautiously than their bigger counterpart due to the fear of potential predators resulting in lower feeding frequency in smaller fishes [30].Larger fish may require more food to obtain the necessary energy for reproductive activity than smaller ones require for growth.In larger fish, a wider mouth opening helps to ingest relatively large quantity food items at a time compared to smaller ones [31].
Quantitative and qualitative changes in fish food during the life span are useful tools to define the diet of a particular fish species [32,33].Planktonic algae were reported to be the dominant food item of L. parsia and planktonic groups according to the order dominance was chlorophyceae, bacillariophyceae and myxophyceae [3] in open environment.The percentage occurrence of filamentous algae (53%) and diatoms (54.1%) was more or less equal through considerable monthly variations were seen among the two food items [34].In the present study,

Feeding ecology, prey selection and growth performances of gold spot mullet, Liza parsia (Hamilton, 1822) in extensive brackish water farming system of Indian Sundarbans
similar order of dominance of planktonic groups was observed in extensive farming system but percentage occurrence of different planktonic groups widely differed from the previous reports.Fish food preferences were determined through electivity index (E) to throw some light on the preferred food of gold spot mullet in extensive farming system.Ribeiro and Nuňer [35] suggested that changes of feeding habits of a fish species are a function of the interactions among several environmental factors that will influence the selection of food item.In the present study, gold spot mullet showed highest preference towards myxophyceae followed by bacillariophyceae, chlorophyceae, copepod, fish and shrimp parts and dinoflagelets.According to Ivlev's equation [36], E varies from -1 to+1, where 0 to +1stands for positive selection, while values between -1 to 0 indicates negative selection of that prey item while true positive or negative prey selection can be interpreted only at values >0.3 or <-0.3 respectively [37].In the present study, L. parsia truly selected myxophyceae though ranking 3rdin the order of dominance in the ambient water making it the prime food material in the stomach.Although being 2nddominant planktonic constituent in water and third in stomach, bacillariophyceae was observed to be the 2nd most preferred truly selected prey group.In spite of being dominant prey group in pond water and 2nd in the order of dominance in the stomach content, chlorophyceae ranked 3rd in order of preference.Chlorophyceae was not truly selected and probably swallowed mechanically during food intake as those were most abundant in water.Gold spot mullet juveniles positively selected copepoda during the initial months of rearing but true negative selection was observed during third month onwards.Fish and shrimp parts and dinoflagelets were not at all selected by L. parsia.Order of preference of phytoplanktonic food by L. parsia as myxophyceae>bacillariophyceae>chlorophyceae is different from coexisting species like Mugil cephalus L bacillario phyceae>myxophyceae>chlorophyceae [29] and Liza tade as ch lorophyceae>myxophyceae>bacillariophyceae [38] in the same environment suggests feeding strategy of mullets to reduce inter specific competition within the same tropic level.
Growth of gold spot mullet in the present study (58.94±0.02g, 15.2±0.19cm) was much higher than 47.68g in 225 days at stocking density of 10000 fish ha-1 as reported by Ali et al [2] and 54.02±2.11gat stocking density of 2000 no's/ ha in 180 days polyculture with M. cephalus (4500 no's/ ha), L. tade (1500 no's/ ha) and P. monodon (20000 no's/ ha) in fed polyculture system [8].L. parsia in the present study showed better survival and DWG compared to those reported by Bhowmik [7].Much lower growth (27.03±2.662g, 12.97 cm) compared to the present study was reported from Bangladesh in shrimp-mullet polyculture [39].Slope value of length-weight relationship in the present study (b=3.028) was much higher than those reported from Southern India as 2.796 and 2.988 for male and female respectively.Better growth and health condition observed in the present study can be attributed to the low production farming system with comparatively lower stocking density providing greater food and space for the growing fishes.

Figure 1 :
Figure 1: Percentage occurrences of suspended food materials in ambient water of extensive farming impoundments selected for gold spot mullet (Liza parsia) feeding ecology study

Figure 2 :
Figure 2: Average percentage occurrences of food materials in gold spot mullet (Liza parsia) stomachs grew in extensive farming system.

Figure 3 :
Figure 3: Monthly percentage occurrences of food materials in gold spot mullet (Liza parsia) stomachs grew in extensive farming system.

Figure 4 :
Figure 4: Electivity index indicating prey preference of gold spot mullet (Liza parsia) grew in extensive farming system

Figure 5 :Figure 6 :
Figure 5: Growth of gold spot mullet (Liza parsia) in extensive farming system (values are expressed as mean ±SE of three studied ponds)

Figure 7 :
Figure 7: Length--weight relationship of gold spot mullet (Liza parsia) grew in extensive farming system

Table 1 :
Numeric percentage of chlorophyceae ranged between 21.03 and 37.41% (29.59±3.01%)during May and April respectively.Among bacillariophyceae, Navicula, Nitzschia, Cyclotella, Basilaria, Diatoma and Melosira were found to be the most abundant genera.Numeric percentage of bacillariophyceae ranged between 8.96 and 30.99% (22.59±4.34%)during July and September, respectively.Anabaena, Nostoc, Spirulina and Oscillatoria was the most dominant genera undermyxophyceae.The percentage composition of myxophyceae varied between 15.06 and 25.92% (21.38±2.33%)throughout the study period.Among zooplankton groups, copepods and dinoflagelats were most dominant while rotifersandcladocera were less abundant.Percentage composition of copepoda ranged between 2.04 to 24.88% and most common Copepod genera was Calanus.Most abundant dinoflagelats genera were Ceratium and Peridinium.Percentage composition of dinoflagelats were recorded minimum during June (4.21%)and maximum during November (20.00%).Water quality parameters of three extensive brackish water ponds selected for gold spot mullet (Liza parsia) feeding ecology study and Tetraedron.

Table 2 :
Feeding intensity in terms of stomach fullness in gold spot mullet (Liza parsia) grew in extensive brackish water ponds