Feed of Siamese fighting fish , Betta splendens , ( Regan , 1910 ) in open pond : live and formulated diets

The growth rate and percent survival of Betta splendens when submitted to formulated diet and live food treatments are evaluated. The three different diets were used and designated as: formulated diet (basal diet); live food diet (plankton) and mixed diet (formulated diet with plankton). The live food diet contained plankton belonging to an open pond. High mortality was reported with live food (plankton) treatment whereas higher percent survival occurred with formulated diet. Highest specific growth rate, weight gain and final weight were reported in the mixed diet treatment and were significantly different (p<0.01) from those in formulated diet and live food treatments. The gut contents of B. splendens in mixed diet and live food treatments comprised, Rotifera and Bacillariophyceae species in high percentages or rather, over 78% of total organisms. Lecane sp. was the most ingested zooplankton species by B. splendens in both treatments (mixed diet and live food), with the phytoplankton species Asterionella sp. and Melosira sp. respectively in mixed diet and in live food, respectively. Results indicated that the formulated diet influenced the water parameters dissolved oxygen, total suspended solids, total dissolved solids and pH. The live food in the open pond was not enough to improve the growth rate and percent survival of B. splendens. The growth performance of B. splendens; had the best results with mixed diet which was capable of maintaining species’s survival (82%) and development in artificial conditions, benefiting the culture management of ornamental fish.


Introduction
The species Betta splendens (Regan, 1910) is one of the best known and popular ornamental fish in the world.The species, native to the Mekong basin, Asia, belongs to the family Osphronemidae and it is one of the most important ornamental fish due to its body coloration, shape, economical and notable commercial value.Success in hatchery production of ornamental fish, regardless of age development for stocking in the grow-out production system is largely dependent on the availability of suitable live food and formulated diet for fingerlings and adults of B. splendens (Thongprajukaew et al., 2011).
Most ornamental fish culture still depends on the supply of live food and formulated diets, both of which crucial for lowering production costs and sustaining commercial activity.Although live food is believed to be the best protein source, but production is costly and sometimes unreliable.In addition it may provide suboptimum nutrition and cause high mortality (Puello-Cruz et al., 2010).
Consequently considerable efforts have been devoted to use a mixed diet in which, live food (plankton) and formulated diet are provided to evaluate the survival and growth performance of ornamental fish.Formulated diet coupled to live food has a comparable nutritional quality and maintain the percent survival and growth rate of ornamental fish at low costs on the global market (Lim et al., 2003).
Live food such as plankton preys has been widely used in fish aquaculture.The open pond has a great and diverse planktonic community that may provide live food to ornamental fish when appropriated methods are used.The planktonic community in the open pond may differ according to the environmental conditions and to its high variability depends on cultivation systems and nutrients input.Current assay used the plankton of the open pond to produce live food for B. splendens.
Many studies have focused on the fact that small live food such as planktonic Rotifera species are required for the development of fry production (Lim et al., 2003;Ogata and Kurokura, 2012).However, several studies reported that fish fed solely on formulated diets, have poor percent survival and low growth rate (Finn and Kapoor, 2008), whereas others researches insist that combination of live food and formulated diets in the early stage may provide higher growth rate and percent survival than live food (Rosenlund et al., 1997;Puello-Cruz et al., 2010).
Current study evaluated formulated diet and live food (plankton) to improve the growth rate and percent survival of 60-days-old B. splendens, during 49-days, with regard to three different diets (formulated diet, mixed diet and live food).It was hypothesized: 1)-the plankton biomass in the open pond was enough to improve growth development and survival of B. splendens; 2)-formulated diet, live food or mixed diet, are more efficient in supporting growth and survival of B. splendens; 3)-the feed treatments influenced the water parameters.

Fish and rearing conditions
The experimental work was performed with Betta splendens juveniles (60-days-old) in a pond at the Aquaculture Center (22°5'S and 47°19'W).The fish were randomly divided in to five plastic boxes (34×54×20 cm), with 15 plastic bottles (2 L) in each box with one fish stocked in each bottle.The experiment was carried out during 49 days.Three different food treatments were provided in each plastic box (Figure 1).The five plastic boxes were suspended 10-cm over the surface, maintaining a 0.34 m water depth in the open pond.The open pond was rectangular in shape and 1.3 deep, with an aerial surface of 70 m 2 in area and volume of 54.6 m 3 .The water supply was obtained from an upstream pond channeled to the open pond by an underground grid tubes.Water renovation (calculated from the discharged volume) was equivalent to 5% of the total pond volume per day.The B. splendens were fed on formulated diet (32% crude protein) at a rate of 3% weight.day - three times a day, except live food (plankton) treatment.The NPK (20-5-20) inorganic fertilizer was applied biweekly (2.2 g.m -2 ) to open pond, while water levels in the open pond were maintained between 1.1 and 1.3 m throughout the period, compensating for seepage and evaporation losses.The initial weight and length of fish in the different treatments were 0.42 ± 0.2 g and 2.64 ± 0.06 cm, respectively.Weight gain and total length of fish were measured individually.Growth performance parameters were as following formulate (Equation 1-4): Condition factor (g cm -3 ) = 100 x (W/L 3 ) (1) where: W = live body weight (g) and L = total body length (cm).Specific growth rate (SGR) was calculated according to Kestemont and Stalmans (1992).

Feed treatments
A basal diet was formulated to contain 32% crude protein and 16,891 J g -1 , gross energy based on feed stuff values reported by Tamaru and Ako (1999).Three different diets were formulated designated as: FD= formulated diet, the water of the bottle from the water supply was not in contact with the water in the open pond; FD+P= mixed diet, formulated diet with plankton from the open pond; P= live food, plankton from the open pond.The bottles with feed treatments FD+P and P had many holes for the circulation of water and plankton.Table 1 shows ingredients and formulation of feed treatments.The ingredients tested were ground in a hammer mill, mixed, moistened and pelleted in a meat grinder.The pellets were dried for 24 hours then fractionated and sieved to obtain a diameter of 0.71-1.00mm.

Environmental aspects
Water and plankton parameters were estimated at 8:00 hours.Samples were collected from a depth of 10 cm.Three samples sites were collected weekly during 49 days of experimental: IW= inlet water; CB= close to boxes and WO= water outlet (Figure 1).Temperature (Temp), pH, conductivity (Cond) and dissolved oxygen (DO) were measured using multi-probe Horiba U-10, and water transparency (Transp) with a Secchi disk.Orthophosphate (OP), total phosphorus (TP), nitrite (NO 2 ) and nitrate (NO 3 ), were quantified with a spectrophotometer according to Golterman et al. (1978) and ammonia (NH 4 ) was quantified as describe by Koroleff (1976).Chlorophyll-a (Chl-a) concentration was determined by colorimetric analysis using a spectrophotometer after extraction with 90% ethanol (Nusch, 1980).Total suspended solids (TSS) and total dissolved solids (TDS) were determined according to Boyd and Tucker (1992).Alkalinity (Alk) and inorganic carbons (free CO 2 -FCO 2 , Bicarbonate -HCO 3 and total  (1978).Analyses were performed immediately after sampling or samples were duly stored under refrigeration.
For zooplankton, 10-L of water was filtered through 58 µm pore net, concentrated to 50 mL, and formalin was added to reach 4% final concentration.Cladocera and Copepoda species were identified in a reticulated acrylic chamber under stereomicroscope (40 × augmentation).Rotifera species were identified and counted in Sedgewick-Rafter chamber under a Leitz microscope (100 × augmentation).For phytoplankton the quantification was estimated by counting the cells using Sedgewick-Rafter chamber during 30 minutes of sedimentation.The density of each species was calculated by counting all specimens totaling at least 100 individuals of the most frequent species or genus (p<0.05;Lund et al., 1958).Plankton concentrations were expressed as number of individuals per m 3 and percentage abundance.Plankton samples were only collected close to the plastic boxes (CB).

Gut contents analysis
At the end of the experiment, ten fish were randomly selected from group that received live food plankton (P) and from the group that received formulated diet with plankton (FD+P), they were kept at low temperature (-5 °C) for 10 minutes (to avoid regurgitation) and subsequently fixed in 10% formalin.Food contents of the whole gut were analyzed quantitatively and qualitatively for their composition.Only food particles that were partially or not digested were analyzed.

Statistical analysis
Principal components analysis (PCA) was used to reduce the dimensionality of the environmental variables and to rank the variation of these variables between sites (Legendre and Legendre, 1998).These analyses were performed using the software Statistica 8.0 (STATSOFT INC., 2007).One-way ANOVA was used for growth performance to compare the effects of the experimental diets.Fisher's test was employed to test differences between average and were regarded significant when p<0.05.Spearman rank correlation (Rs) coefficient analysis was applied to compare to fish feeding (Siegel, 1975).This analysis was employed to study correlations between the available plankton in open pond and gut content of two feed treatments formulated diet with plankton (FD+P) and plankton (P).

Results
Specific growth rate, weigh gain and final weight were higher (p<0.01) in FD+P treatment when compared with live food and formulated diet (Table 2).Fish showed lower specific growth rate, weigh gain and final weight in the live food (P) treatment.No significant differences (p>0.05) for condition factor were reported between formulated diet, live food and mixed diet feed treatments.Fish showed higher total length with formulated diet (FD) and mixed diet (FD+P) feed treatments, which were significantly different with live food (P) treatment (p<0.01).Fish survival was lower in P feed treatment (44%), whilst the highest percent survival (92%) was recorded in FD feed treatment, which was different from percent survival (80%) of FD+P feed treatment (Table 2).
Data obtained from water analysis were submitted to the principal component analysis (PCA) and results (open pond) are given in Figure 2. The first axis (PC1), accounting for 66% of the data variability is bipolar with three variables in its positive side (TDS, DO and pH) and, more eutrophic was observed at the WO (water outlet) site associated with chlorophyll-a, alkalinity, temperature, conductivity, inorganic carbon, ammonia, nitrite, phosphorus, orthophosphate and transparency, with highest values.The second axis (PC2) accounting for a further 34% of the data variability, was associated only nitrate at a negative coefficient and related to IW site.Analyses of sites showed the formation of three different groups (Figure 2).
Since temperature was approximated 24 °C during the experiment and pH was too high (>7.0),alkalinity (mean 49 mg.L -1 ) and bicarbonate (mean 60 mg.L -1 ) were high in the inlet water.Conductivity, chlorophyll-a, orthophosphate and total phosphorus had rather homogeneous values with peak in the second experimental week (7-days) after which they decreased until the end of the experiment.Dissolved oxygen (4.4-6.0 mg.L -1 ) was low in the 7-day experiment but latter increased until the end of the experiment (Figure 3).
No significant difference (p>0.05) was registered between feed treatments FD+P (R s = 0.42) and P (R s = 0.50) for gut contents of B. splendens (Table 3).The food organisms eaten by B. splendes were diverse and included phytoplankton.Both feed treatments were accepted by the ornamental fish.High ingestion of Melosira sp, Bacillariophyceae taxon, was probably due to its great abundance (17%) in the open pond when, compared to that of others species, such as Asterionella sp and A. Formosa, Bacillariophyceae taxon, and Rotifera species such as Proales sp., Lecane sp., L. bulla, Colurella uncinata and Cephalodella sp (Table 3).Asterionella sp (FD+P) feed treatment and Melosira sp (P) feed treatment, represented more than 37% and 35% of phytoplankton ingested by ornamental fish, respectively.The Copepoda species had a higher abundance in the open pond but represented less than 16% in gut content, contrastingly to Proales sp and Lecane sp (Rotifera) which represented more than 28% of zooplankton ingested by B. splendens.A high amount of resting eggs in the gut content was observed.Cyanobacteria species were observed only in the water column in the open pond (Table 3).

Discussion
Betta splendens did not select (p>0.01)planktonic species in the treatments, and Bacillariophyceae and Rotifera were more representative in the gut contents.The same values of plankton was recorded in B. splendens fed on live food (P) and mixed diet (FD+P) but the highest growth performance was recorded in mixed diet and indicated that the association of formulated diet and plankton was enough for fish metabolism which consequently resulted in a more suitable growth, than live food only or formulated diet only.The highest occurrence of algae and Rotifera in gut content revealed the planktivorous nature of feeding of B. splendens in the open pond.
Betta splendens fed on mixed diet (75% formulated diet and 25% live food) registered the highest specific growth rate, weigh gain and condition factor.Mandal et al. ( 2010) observed that 25% replacement of live food with formulated diet was better for a higher growth rate of ornamental fish.These results agree with current study that the highest growth performance was obtained with FD+P feed treatment, but the highest percent survival occurred with FD feed treatment (90%).
The low percent survival of B. splendens in P treatment is probably due to the high density of Copepoda (72.37%), and the most these species have a rapid jumping movement, live preys may be less available to the fish for predation than Rotifera that represented 18.86% of the total zooplankton organisms in the open pond.Olivotto et al.
(2010) demonstrated high growth rate and percent survival of Amphiprion clarkii, an ornamental fish, when a combined diet was used.Fish would spend less time and energy to ingest rapidly available artificial food, than to ingest the same volume of plankton by sucking water (Rahman et al., 2010).This argument explains the higher growth rate and percent survival of B. splendens in mixed diet (FD+P) and formulated diet (FD).
The feed treatments which influenced the abiotic properties of water and transparency were higher at WO site.Highest nitrogen compounds in the three sites were due to water supply.Highest amount of suspended materials (TSS and TDS) occurred close to the boxes and were caused by the decomposition of organic matter (feed waste).The constant water flow through the open pond carried the suspended material and moving feed waste to the WO site, causing increased of nutrients (TP, OP and NH 4 ), phytoplankton biomass (Chl-a), carbon contents and water conductivity, which indicated an increase in eutrophication.No pH variation occurred in the open pond at different sites, possibly to photosynthesis.Results indicate that feeding not affect pH values.Similar observations were reported by Hassan (2011) on the common carp.
Results from the current study indicated that feed treatments influenced some water parameters such as TSS, TDS, DO and pH, whereas live food (P) in the open pond was not enough to improve the growth rate and percent survival of B. splendens.The growth and survival of B. splendens, registered their best results with mixed diet (FD+P), but the highest percent survival was obtained with formulated diet (FD).Hence, concluded mixed diet is a good tool for successfully in, maintaining the percent survival (82%) and development of B. splendens in artificial conditions, for the benefit of the culture management of this ornamental fish.

Figure 1 .
Figure 1.Geographic location (1), layout of Aquaculture Center (2) and outline of the flow-through open pond (3) used with the five plastic boxes and 15 plastic bottles (2 L) in each box, where: ο= FD+P feed treatment (mixed diet); ▲= FD feed treatment (formulated diet = basal diet) and •= P feed treatment (live food = plankton).

Figure 2 .
Figure 2. Interpolation of eigenvalues from the matrix of water variables from the open pond.The first two axes from the principal component analysis (PCA), where: IW = inlet water; CB = close boxes; WO = water outlet; open circle = water variables; closed triangles = sites.

Table 1 .
Formulation and proximate composition of the formulated diet (basal diet) for rearing Betta splendens juveniles in 49-days experiment.Betta splendens in open pond CO 2 -TCO 2 ) were determined according to Mackereth et al.

Table 2 .
Biological indices obtained for Betta spleendens juveniles fed on formulated diet (basal diet -FD), live food (plankton -P) and mixed diet (formulate diet with plankton -FD + P) in 49-days experiment.Values in the same row with different superscripts are significantly different (p<0.05).

Table 3 .
Relative abundance (%) and number of individuals.m 3 × 10 4 (Ind.),groupedbytaxa found in the gut contents of Betta splendens in feed treatments(FD+P and P)and in the open pond.