Effect of partially replacing fish meal with honey locust seed (Gleditsia triacanthos) on growth, feed utilisation and body composition of mirror carp, Cyprinus carpio, fingerlings

Abstract

Honey locust seed (HLS) (Gleditsia triacanthos) was evaluated as a partial replacement for fish meal (FM) in Cyprinus carpio fingerlings diets formulated to contain 38% crude protein, 13% lipid and 19.9 MJ/kg gross energy. An 60-day feeding trial was conducted in a recirculation system aquarium. HLS was included in the diets at different levels (0, 100, 200, 300 and 400 g raw and 100, 200, 300 and 400 g heat-treated for diets Control, Rw₁, Rw₂, Rw₃, Rw₄ Ht₁, Ht₂, Ht₃ and Ht₄, respectively) and FM was replaced by 0%, 5%, 10%, 15% and 20% of HLS (raw and heat-treated). The fish were handfed daily at of 3% of their body weight during the experiment. Diet had a significant effect on the growth performance. Heat treatment increased the inclusion level of HLS. There were no significant differences (P>0.05) among fish fed diets Control, Rw₁, Rw_2, HT₁, HT₂ and HT₃ in terms of growth performance. Fish fed diets containing higher levels (>30%) of raw HLS seed had significantly lower whole body fat content. Raw and heat-treated HLS has potential as an alternative feed ingredient. These results show that up to 20% Raw HLS can be used to replace 10% of FM in diets for mirror carp. However, HLS should be heat-treated if inclusion rates are to exceed 20% of the diet.

Keywords:

• Honey locust
• Gleditsia triacanthos
• mirror carp
• Cyprinus carpio
• growth

Introduction

Reducing FM level is key to reducing feed cost for commercial fish farming and ensuring sustainability of this enterprise. It is essential to evaluate the suitability of alternate plant and animal protein meals as dietary protein sources for fish species (Yan et al. 2006). Locally available plants are often less expensive sources of energy and protein for commercial feed formulations. Meals made from peas are intermediate in terms of both a good energy source and a reasonable amount of protein to the ration. However, inclusion levels in feed formulations must be tempered by cost, processing consideration, nutrient availability as well as palatability of the meal to target species. As a result, information related to nutrient availability and the biological response of the target species is essential to the successful utilisation of a given ingredient (Davis et al. 2002). Recently, considerable emphasis has been centred on the use of unconventional protein sources, from plant products proteins in particular towards those not directly usable for human consumption (Kaushik et al. 2004; Buyukcapar and Kamalak 2006; Yue and Zhou 2008; Yanar et al. 2010). These studies have now focused on utilising less expensive and readily available plant protein sources to replace FM, without reducing the nutritional quality of feed.

Gleditsia triacanthos, known as ‘Honey locust’, is a legume crop widely available in all over the Mediterranean area. Although Honey locust was collected from natural pasture in some parts of Turkey and used for sheep and goat feeding by some local farmers, there are no available data on commercial production of Honey locust in Turkey. To date, there have been no published studies on the use of Honey locust seed (HLS) as protein source in the diet of mirror carp (Cyprinus carpio) fingerlings.

This study was conducted to evaluate growth performance of C. carpio diets in which FM was gradually replaced by HLS and to evaluate the effect of replacing FM with HLS on body composition of mirror carp.

Materials and methods

Mirror Carp (C. carpio) fingerlings were obtained from the brood stock maintained at the State Hydraulic Works, Adana, Turkey. The fish (about 10 g) were transferred to the Department of Fisheries, Faculty of Agriculture and University, Kahramanmaras Sutcu, Imam and acclimatised for 10 days in a 250-L fibreglass tank at 27.8±0.4°C. During this period, they were fed with a diet containing approximately 350 g/kg protein, 80 g/kg lipid, 7 g/kg ash and energy content of 13.54 MJ/kg dry matter until they reached a body weight of approximately 13 g.

Fish fingerlings were randomly allocated at a stocking rate of 10 fish per aquaria with three replicate aquariums under each dietary treatment. The fish were hand fed twice a day at 8.30 am and 16.30 pm at 3% of their body weight during the experimental period of 60 days. They were individually weighed at the beginning of the experiment (day 0) and at day 20, 40 and 60. At the beginning of the experiment 10 fish from the same stock were sampled to determine the whole body composition. At the end of the experiment, five fish per aquarium were sampled to determine the whole body composition in groups. HLS was included in the diets at different levels (0, 100, 200, 300 and 400 g raw and 100, 200, 300 and 400 g heat treated for diets Control, Rw₁, Rw₂, Rw₃, Rw₄, Ht₁, Ht_2, Ht₃ and Ht₄, respectively) and FM was replaced by 0%, 5%, 10%, 15% and 20% of HLS (raw and heat-treated). The growth parameters of fish fed these diets were compared to fish fed a FM and soybean meal based control diet. The control diet contained 40% FM and 15% of soybean meal as main protein sources.

A recirculation system containing a set of aquaria, each with a capacity of approximately 80 L and water depth 50 cm. Water quality was monitored through the experiment. Temperature, pH and dissolved oxygen were recorded daily. Mean temperature, pH and dissolved oxygen were 28.1±0.04°C, 7.8±0.08 and 6.6±0.04 mg/L, respectively. Water quality was controlled by biologic filter and electronic heater. Water was re-circulated through the filter at a rate of 5 L/min. Additional aeration was provided by an air pump.

HLS was sun dried and ground into powder to pass a 1 mm sieve. Ground HLS was heat-treated in autoclave at 121°C for 10 min to eliminate the possible detrimental effect of anti-nutritional factors (Rehman and Salariya 2005). The chemical composition of FM, soybean meal, raw and heat-treated HLS are given in Table 1.

Table 1. Proximate composition (g/kg) and estimated gross energy (MJ/kg) contents of fish meal (FM), soybean meal (SBM), maize flour (MF) and Honey locust seed (HLS) used in the experimental diets.

	Feedstuffs
Constituents	FM	SBM	MF	HLS (Raw)	HLS (heat-treated)
Dry matter	916.6	890.0	881.3	922.2	931.2
Crude protein	663.6	470.0	78.0	210.1	200.9
Crude fat	109.3	10.0	35.2	23.0	22.8
Crude ash	125.0	60.0	55.2	44.2	45.2
Crude fibre	5.0	60.0	26.2	89.0	86.3
Gross energy^a	20.3	17.5	16.21	16.36	16.73
Anti-nutritional factor
Condensed tannin	ND	2.6	0.7	3.7	2.3
Note: Crude protein, crude lipid, ash and gross energy content are expressed on a dry matter basis. ND, not determined.
^aGross energy was estimated using the following coefficients: 23.6 kJ/g for crude protein; 39.5 kJ/g for crude fat and 17.2 kJ/g for carbohydrates (NRC 1993).

Growth performances were evaluated by body weight gain (BWG), feed conversion ratio (FCR), feed intake (g) per fish (FI), specific growth rate (SGR%) and protein efficiency ratio (PER).

In the feeding experiment FM was replaced with raw and heat-treated HLS on a dry matter basis to maintain crude protein and energy levels. The composition of the experimental diets is given in Table 2. Ingredients were mixed, formed into a moist pellet about 2 mm in diameter and then dried in a forced-air oven at 45°C. Nine isonitrogenous and isoenergetic diets were formulated to evaluate nutrition value of HLS for carp fingerling (Table 2). HLS, experimental diets and fish samples were analysed for their proximate composition according to the methods of AOAC (1990). Table 2 shows the formulation, proximate composition of the nine diets. Condensed tannin was determined by butanol–HCl method as described by Makkar et al. (1995).

Table 2. Ingredients and chemical compositions of the experimental diets (as fed).

	Diets
	Control	Rw1	Rw2	Rw3	Rw4	Ht1	Ht2	Ht3	Ht4
Ingredients (g/kg)
Fish meal	400	380	360	340	320	380	360	340	320
Soybean meal	150	150	150	150	150	150	150	150	150
HLS	0	100	200	300	400	100	200	300	400
Maize flour	367	284	202	122	42	284	202	122	42
Sunflower oil	72	75	77	77	77	75	77	77	77
DCP^a	2	2	2	2	2	2	2	2	2
Vitamin–Minarals^b	5	5	5	5	5	5	5	5	5
Methionine	2	2	2	2	2	2	2	2	2
Limestone	1	1	1	1	1	1	1	1	1
Salt	1	1	1	1	1	1	1	1	1
Analysed proximate composition (g/kg)
Dry matter	922.3	918.4	920.2	919.3	918.3	921.3	919.5	918.5	917.6
Crude protein	380.2	380.6	381.2	382.2	382.6	379.2	379.5	380.3	379.6
Crude fat	131.6	131.7	130.9	128.1	125.4	130.9	131.3	128.9	127.3
Crude ash	65.2	62.5	66.2	64.5	63.4	62.1	64.8	63.2	61.3
Crude fibre	40.3	45.5	44.7	47.1	49.5	42.2	44.2	46.3	48.4
Gross energy MJ/kg^c	19.9	19.8	19.9	19.8	19.9	19.8	19.9	19.8	19.9
Anti-nutritional factor (g/kg)
Condensed tannin^3	0.6	0.9	1.2	1.5	1.8	0.8	1.0	1.2	1.4
Note: Diet Rw1, Rw2, Rw3, Rw4 containing 100, 200, 300 and 400 g Raw HLS, respectively, Diet Ht1, Ht2, Ht3 and Ht4 containing 100, 200, 300 and 400g heat-treated HLS, respectively.
^aDi calcium Phosphate.
^bPer 5 kg vitamin-mineral premix: 20.000.000 IU vitamin A, 2.00.000 IU vitamin D3, 200.000 mg vitamin E, 12.000 mg vitamin K3, 20.000 mg vitamin B1, 30.000 mg vitamin B2, 200.000 mg niacin, 50.000 mg Ca-panthothenate, 20.000 mg vitamin B6, 50 mg vitamin B12, 500 mg D-biotin, 1.200 mg folic acid, 200.000 mg vitamin C and 300.000 mg inositol.1.200.000 mg cholin chloride, 40.000 mg manganese, 30.000 mg Zinc, 800 mg Copper, 1.000 mg iodine, 150 mg Selenium, 40.000 mg magnesium.
^cCalculated values.

A completely randomised design was adopted with three replicates per diets. One-way analysis of variance (ANOVA) was carried out to determine the effect of diets on growth and whole body composition parameters using General Linear Model (GLM) of Statistical for windows (1993). Significance between individual means was identified using the Duncan's multiple comparison tests. Mean differences were considered significant at P<0.05.

Results and discussion

Fish fed diet RW₂ containing 20% raw HLS had growth performance similar to those fish fed with the control diet. There was inverse relationship between the growth of mirror carp fingerlings and dietary HLS levels beyond 20% inclusion level. This could be the result of several factors, including reduced nutrient and energy digestibility and energy utilisation efficiency. The FI and PER also decreased and FCR increased with increasing inclusion level of raw HLS.

Diets containing higher than 20% raw HLS a significantly poorer growth and feed utilisation compared to the control and HT₁, HT₂ and HT₃ diets which contained heat-treated HLS. It is likely that the anti-nutritional factor levels such as condensed tannin in diet RW₂ were not high enough to affect the digestibility of protein and energy. The reduced growth performance of fish fed other diets might be related to the presence of various anti-nutritional factors in HLS. HLS contains considerable amount of condensed tannins (Table 1). Buyukcapar and Kamalak (2006) showed that fish fed diets containing 10% of raw and 10%, 20% and 30% heated Vicia peregrina seed had a growth performance similar to those of the control diet while fish fed diets containing >10% raw Vicia peregrine seed had significantly lower growth performance than the control diet. This observation is also consistent with the findings of Borlongan et al. (2003) and Hossain et al. (2002).

However, the current experiment clearly showed that the inclusion level of heat-treated HLS into diet of mirror carp fingerlings could be increased (Table 3). The growth performance of mirror carp fingerlings fed diets containing up to 20% of heat-treated HLS was comparable with control diets (Table 3) possible due to reduction in condensed tannin in diets by heat treatment (Table 2). Makkar et al. (1987) indicated that the presence of tannin and trypsin growth inhibition were related to their lower nutritive. Although peas are considered to contain relatively low levels of anti-nutritive factors, performance of the fish could nevertheless be impaired by them (Francis et al. 2001). Extrusion decreases the efficiency of the majority of these (Rumsey et al. 1993). Protein and lipid content of fishes fed diets RW₃, RW₄ and HT₄ were lesser as compared to control diet fed fishes (Table 4). The body lipid and protein contents values were decreased with increasing incorporation of HLS in the diets. This agrees with the result of Poumogne et al. (1997) and Siddhuraju and Becker (2001) on juvenile Nile tilapia and C. carpio, respectively, where a similar reduction in body lipid occurred in fish fed diets containing plant-based proteins. The causative factors here may have been a reduction in fishmeal and thus decreased availability of lipid contributed by fishmeal in the diet, and the fact that the crude lipid contributed by HLS contained anti-nutritive factors. This result is in agreement with findings of Buyukcapar and Kamalak (2006) who observed similar reductions in whole body in mirror carp fed diets containing Vicia peregrine seed.

Table 3. Growth parameters of mirror carp obtained in experiment (n=30).

	Diets
Parameters	Control	Rw1	Rw2	Rw3	Rw4	Ht1	Ht2	Ht3	Ht4
In W (g)	13.2^a±0.35	13.1^a±0.34	13.5^a±0.32	13.5^a±0.28	13.2^a±0.25	13.1^a±0.32	13.5^a±0.30	13.3^a±0.35	13.4^a±0.29
Fi W (g)	33.8^a±0.97	33.6^a±1.02	33.9^a±1.03	30.1^b±1.05	20.3^c±1.11	33.8^a±1.10	33.6^a±1.13	32.2^ab±1.09	29.5^b±0.73
BWG (g)	20.5^a±0.52	20.4^a±0.46	20.3^a±0.66	15.5^c±0.29	7.1^d±0.19	20.6^a±0.58	20.1^ab±0.58	18.9^b±0.47	16.1^c±0.47
FI (g)	35.0^a±0.05	34.3^a±0.49	33.9^ab±0.45	33.3^b±0.41	27.7^d±0.26	34.9^a±0.32	34.3^ab±0.20	33.6^b±0.46	31.9^c±0.42
FCR	1.7^a±0.15	1.6^a±0.08	1.6^a±0.07	1.9^a±0.11	4.0^b±0.22	1.7^a±0.18	1.7^a±0.10	1.7^a±0.09	2.0^a±0.1
SGR%	1.55^ab±0.1	1.56^ab±0.1	1.53^ab±0.1	1.33^ab±0.1	0.71^c±0.1	1.57^a±0.1	1.52^ab±0.1	1.47^ab±0.1	1.31^b±0.10
PER	1.61^ab±0.13	1.61^ab±0.08	1.57^ab±0.07	1.34^b±0.06	0.67^c±0.04	1.63^a±0.14	1.56^ab±0.09	1.5^ab±0.07	1.34^b±0.06
^a–cRow means with common superscript letters do not differ (P>0.05), Diet Rw1, Rw2, Rw3, Rw4 containing 100, 200, 300 and 400 g raw HLS, respectively, Diet Ht1, Ht2, Ht3 and Ht4 containing 100, 200, 300 and 400 g heat-treated HLS, respectively. Control diet not containing HLS, In W: initial body weight (g), Fi W: final body weight, BWG: body weight gain, FCR: feed conversion ratio, FI: feed intake (g/fish), SGR: specific growth rate (%), PER: protein efficiency ratio.

Table 4. The whole body composition analyses (%) (on wet weight basis) of mirror carp fed test diets (n= 15).

	Diets
%	Initial	Cont.	Rw1	Rw2	Rw3	Rw4	Ht1	Ht2	Ht3	Ht4
DM	18.3±0.2	25.2^a±0.4	25.4^a±0.5	24.9^a ±0.3	24.8^a±0.3	24.7^a±0.3	25.5^a ±0.6	24.4^a±0.2	24.9^a±0.3	24.8^a±0.2
Protein	11.4±0.3	15.5^ab±0.3	16.0^ab±0.3	16.3^a±0.3	15.0^bc±0.3	14.6^c±0.4	15.8^ab ±0.4	15.9^ab±0.3	15.8^ab ±0.1	14.5^c±0.4
Fat	4.2±0.1	7.2^a±0.3	6.7^a±0.2	7.1^a±0.2	6.9^ab±0.2	4.8^c±0.2	7.0^a ±0.4	6.7^a±0.3	6.6^a ±0.2	5.7^b±0.2
Ash	2.2±0.1	2.6^a±0.2	2.5^a±0.1	2.4^a±0.2	2.5^a±0.2	2.4^a±0.3	2.4^a±0.1	2.4^a±0.1	2.5^a±0.2	2.4^a±0.1
^a–dRow means with common superscript letters do not differ (P>0.05), DM: dry matter, Diet Rw1, Rw2, Rw3, Rw4 containing 100, 200, 300 and 400 g raw HLS respectively, Diet Ht1, Ht2, Ht3 and Ht4 containing 100, 200, 300 and 400 g heat-treated HLS, respectively. Control diet not containing.

G. triacanthos seed has potential as an alternative feed ingredient and can be used up to 20% of diet as protein source in diets of mirror carp fingerling without any adverse effects. G. triacanthos seed should be heat-treated if inclusion rate exceeded 20% of diet. It is likely that the use of locally available and relatively cheap G. triacanthos seed would reduce the feeding cost considerably as an alternative protein sources for fingerling mirror carp.

Acknowledgements

The authors would like to thank the Resource Fund of the University of KSU (Turkey) for their financial support (2009/3–12) of the experiment.