MATERIALS AND METHODS

Factorial experiment, with two replications for each of the treatments, was conducted under ambient condition using earthen ponds. After preliminary preparations (Javed, 1988), all the ponds were initially fertilized, separately, with 40 kg broiler droppings (3333.33 kg ha^–1) as a starter dose to stimulate primary productivity. Fingerling major carps, 6-7 months old (induced bred, procured from Fish Seeds Hatchery, Faisalabad), average weight 21.32 ± 1.99 g, were randomly stocked, from a selected population, in each of the ponds with stocking density of 25, 60 and 15% for Catla catla, Labeo rohita and Cirrhina mrigala respectively (64 fish in each of the ponds). Fertilization of ponds with broiler droppings (4.37 % nitrogen) was started on the basis of nitrogen contents. Five levels, viz. 0.10, 0.13, 0.16, 0.19 and 0.22 g nitrogen per 100 g of wet fish weight daily, were used as nitrogen treatments. However, sixth treatment served as control (without additives). For the quantitative and qualitative study of plankton, from each of the five sub-stations at each pond, two samples were collected both from the surface and from the bottom. The method of microscopic examination as described in APHA (1975) was employed following the sand filtration procedure for the enumeration of phytoplankton (Boyd, 1981). Zooplankton, insect larvae and other animals were studied by taking 10 liters of each pond water from surface, column and bottom at each sub-station. These were pooled and filtered through a plankton net fitted with a glass bottle of 400 ml capacity (mesh size 56 μ). The retenate containing different animals was preserved in Longyi acid and stored in refrigerator. A 5 ml portion of the above sample was taken for the quantitative and qualitative study of zooplankton. The organisms were identified under the microscope upto generic level. The identification of fauna and flora was made by using the books (literature) of Ward and Whipple (1959), Desikachary (1959), Hegner and Engemann (1968) and Marshall and Williams (1972). Data were analyzed for ANOVA and DMR tests. Correlation and regression analyses were also performed to find out relationships/trends among various parameters under study.

RESULTS

Table 1 shows the mean phytoplankton densities in ponds fertilized with broiler droppings at five nitrogen levels and control (without additives).

Table 1:	Mean (±SD) phytoplankton productivities (Nos./5 ml of water/month) in fish ponds
Means with similar letters in a single row are statistically similar at P<0.05.

Table 2:	Mean (±SD) zooplankton productivities (Nos./5 ml of water/month) in fish ponds
Means with similar letters in a single row are statistically similar at P<0.05.

Table 3:	Mean planktonic productivities of ponds under different levels of nitrogen
Column means with similar letters are statistically similar at P<0.05: N = Nitrogen; x = Independent variability; Y = Dependent variable; Nos = Numbers.

Table 4:	Regression of increase in fish yield on phytoplankton and zooplankton productivities
x = Independent variable; Y = Dependent variable; Nos = Numbers: N = Nitrogen

Six genera of Chlorophyceae, viz. Chlamydomonas, Closterium, Microspora, Oedogonium, Pandorina, Spirogyra; two genera of Chrysophyceae including Botryococcus, Synura; and three genera of Bacillariophyceae (Cyclotella, Cymbella and Navicula) were recorded during different months in six treatments. As regards Dinophyceae and Euglenophyceae, the genera observed were Peridinium and Euglena respectively. Myxophyceae included the genera Anabaena, Microcystis and Oscillatoria.

Among phytoplankton Chlamydomonas, Microspora, and Spirogyra showed significantly maximum mean densities under 0.10 g N level while under 0.13g N level Pandorina and Botryococcus showed significantly maximum mean densities than rest of the treatments. Four genera viz. Chlamydomonas, Spirogyra, Navicula and Peridinium showed significantly higher densities in pond water under 0.16 g N level. However, 0.19 g N level promoted significantly higher densities of genera Pandorina, Synura, Cyclotella, Cymbella, Navicula, Euglena, Anabaena, Microcystis and Oscillatoria. However the densities of Pandorina under 0.13 and 0.19 g N levels were statistically non-significant. Navicula densities under 0.16 and 0.19g N levels were statistically non-significant also. Under 0.22 g N level the densities of Closterium, Oedogonium, Botryococcus, Cyclotella, Euglena and Nostoc were statistically higher than rest of the treatments. However, Closrerium, Cyclotella and Euglena showed non-significant differences between 0.19 and 0.22 g N levels. The control treatment exhibited significantly lower densities of phytoplankton than all the five treatment levels (Table 1). Table 2 shows the mean of zooplankton densities in ponds. Zooplankton represented by ciliates (Protozoans) and 10 other genera in all treatments except under 0.16 g N level 112 genera). The genera belonged to Phyla, viz. Rotifera and Arthropoda (crustaceans), were the inhabitants of different treatments. Zooplankton viz. Bosmina and Diaptomus showed significantly higher densities under 0.10 g N level than rest of the treatments. Under 0.13 g N level Ciliates, Branchionus, Bosmina, Cyclops, Daphnia, Diaptomus and Moine showed significantly high distribution. 0.16 g N level promoted significantly high densities of Asplanchna, Branchionus, Keratella, Mytilina, Polyarthra, Cantliocamptus, Cyclops, Cypretta, Daphnia and Moine. Ciliates and genera Asplanchna, Mytilina, Bosmina, Cyclops, Daphnia, Diaptomus and Moines showed significant distribution under 0.19 g N level. However, 0.22 g N level provided suitable environment for the significantly high occurrence of Ciliates, Bosmina, Cyclops and Diaptomus. The response of control treatment towards Zooplankton productivity was significantly lower than the five treatments. Insect larvae, nymphs and tadpoles of frog showed significantly variable occurrence.. under different treatments also (Table 2).

Mean annual phytoplankton productivity under 0.19 g leve of nitrogen was the highest followed by the productivities under 0.22, 0.16 and 0.13 g levels of nitrogen with statistically non-significant differences. However, the same under control treatment was the lowest with the value of 22.75 individuals per 5 ml of water (Table 3). Zooplankton productivity was the best under 0.16 g level of nitrogen (63.83 individuals per 5 ml of water). However, the productivity under this level showed non-significant difference with 0.19 g level of nitrogen. Increasing the level of nitrogen showed significant increase in the zooplankton productivity upto 0.16 g level of nitrogen. However, the increase beyond 0.16 g nitrogen level showed gradual decrease in zooplankton productivity upto 0.22 g nitrogen level. The correlation coefficients between phytoplanktor and zooplankton productivities, under all the treatments were positive and significant (Table 3).

Table 4 shows the regression of increase in fish yield o the phytoplankton and zooplankton productivities of ponds under six treatments. Under all the treatments increase in fish yield had positive and highly significant regression on phytoplankton productivity except for control treatment (p< 0.05). Increase in fish yield also showed positive and highly significant regression on zooplankton productivities under all the treatments except for 0.22 g N level (p<0.05).

DISCUSSION

As regarded zooplankton productivity of ponds, both 0.16 and 0.19 g N levels responded equally well while the third best treatment for zooplankton productivity was 0.22 g N level (Table 3), however, control responded poorly for zooplankton growth. The correlation coefficients between phytoplankton and zooplankton densities were positive and significant for all the treatments (Table 3). Javed et al. (1995) reported direct correlation between phytoplankton and zooplankton productivities in major carps rearing ponds under broiler droppings, cow-dung and layer droppings fertilized ponds (added at the rate of 0.10g nitrogen/100 g of fish weight daily). Khan and Siddiqui (1976) reported direct correlation (r = 0.98) between chlorophyll-a content and phytoplankton. However, the correlation between zooplankton and phytoplankton was negatively significant. Several possible explanations may account for the prominent relationships between zooplankton and phytoplankton in a pond ecosystem because an actively grazing Diaptomus may reduce the standing crops of algae (Hazelwood and Parker, 1961), such an activity would produce a negative correlation between zooplankton and phytoplankton. This type of negative correlation could be observed as the grazing effect of zooplankton upon phytoplankton (Sladecek, 1958). But the positively significant correlation between phytoplankton and zooplankton, as observed during this investigation, was due to the responses of different treatments for successive production of phytoplankton which were significantly more than were used either by the fish or zooplankton. Thus, the specific phytoplankton and zooplankton productivity indices of fish ponds may depict the responses of different treatments towards fish yield increments in an integrated semi-intensive major carps polycultrure systems.