Dried duckweed and commercial feed promote adequate growth performance of tilapia fi ngerlings

The present study evaluated Nile tilapia when fed on dried duckweed in combination with commercial feed, aiming at reducing production costs. Three diets, which consisted of commercial feed (40% crude protein), dried duckweed or a combination of commercial feed and dried duckweed were fed to triplicate groups of 20 tilapia (Oreochromis niloticus) fi ngerlings. Fish weighing 3.2 ± 0.94g were stocked in nine 1-m3 net cages and fed twice a day for 50 days. The fi nal average weights of fi ngerlings fed commercial feed (21.67g) and 50% feed + 50% dried duckweed (19.53g) were not different (p<0.05). Likewise, the specifi c growth rate did not differ between fi sh receiving these dietary treatments. Weight gain decreased as water temperature decreased along the experimental period, increasing feed conversion mainly for fi ngerlings fed on dried duckweed only. Also, when fed alone dried duckweed promoted the lowest weight gain and specifi c growth rate. It was confi rmed that dried duckweed can replace 50% (1:1 dry weight) tilapia fi ngerling 40% crude protein feed for a period of at least 50 days, without reducing growth.


Introduction
In Brazil, Nile tilapia (Oreochromis niloticus) is cultivated in semi-intensive production systems in fertilized ponds or intensive systems such as "raceways" and cages. Depending on the culture system adopted, feed can represent 40 to 70% of the total production costs (Kubitza, 2000). However, due to its omnivorous feeding habit, tilapia can tolerate the inclusion of plant protein sources in their diet, thus reducing feed expenses. In Southern Brazil, Nile tilapia is raised in polyculture with Chinese carp (80% tilapia 20% carp) and integrated to swine production to optimize feeding at different trophic levels in the pond. This production system is adopted mainly by small farmers and has a low production cost since tilapia does not receive commercial feed until it reaches 100 to 150g body weight. Pond primary production is responsible for tilapia growth up to that size (Souza Filho et al., 2003).
The macrophytes of the Lemnaceae family, known as duckweeds, are the smallest fl owering plants in the world. Duckweed meal has been used for cattle, poultry, swine and fish feeding, showing favorable results (Skilicorn et al., 1993). Duckweed productivity can reach from 10 to 30ton dry matter/ha/year, depending on the species, climatic conditions, available surface area, amount of nutrients and management. Duckweed protein production/ha is higher than that of most known vegetable crops and twice as high as that of soybeans (Gijzen and Khondker, 1997), assuming an annual production of 17.6ton/ha/year (as dry matter), with a crude protein content of 37% (as dry matter). Fasakin et al. (1999) observed that the inclusion of up to 30% duckweed meal in tilapia diets promoted adequate growth. Hasan and Edwards (1992) evaluated duckweed as a feed item for Nile tilapia under different feeding rates (from 1 to 6% body weight) and observed that the optimum daily feeding rates of Lemna were 5, 4 and 3% (as dry matter) for fi sh of 25 to 44g, 45 to 74g and 75 to 105g in weight, respectively. Mohedano et al. (2005) obtained a reduction of 30% in tilapia feeding costs by substituting duckweed meal for fi sh meal for Nile tilapia. Although the nutritional value of duckweed for tilapia is well known (Iqbal, 1999), the manufacture of duckweed meal to include as an ingredient in commercial feed could be troublesome. Therefore, in the present study we investigated whether dried duckweed alone or in combination with commercial feed could be offered directly to Nile tilapia, without affecting growth performance.

Experimental design
The feeding trial was conducted in nine 1-m 3 net cages, housed in a 100-m 2 earthen pond, with water fl ow of approximately 50L/min. Net cages were set on bamboo stales and attached to wooden structures. Feeding rings (50-cm diameter) were placed in each cage to avoid feed mixing among experimental units.
Nile tilapia fingerlings were obtained from a commercial fi sh farm and were acclimated in 4-m 3 net cages before starting the feeding trial. From this group, 180 fi sh (average weight 3.2 ± 0.94g) were randomly distributed into nine 1-m 3 net cages. A completely randomized design was adopted, where three diets were fed to triplicate groups of 20 fi ngerlings.
Dried duckweed as an alternative dietary source for tilapia

Duckweed production
A small portion of duckweed was cultivated in two 1000-L plastic tanks to reach 2-kg biomass (as dry matter). Swine manure was added weekly as a fertilizer (1 kg/tank). The total duckweed biomass produced was dried and cut to a smaller size to facilitate feeding management. During the growth period, three samples of duckweed were collected from the surface of each tank with a plastic screen, dried at 50°C for 4h, stored in plastic bags, and cooled (-4°C) until proximate composition analyses were performed.

Experimental diets
The diets tested were: 40% crude protein commercial feed, dried duckweed, or 50% dried duckweed + 50% commercial feed. Proximate compositions of the diets are shown in table 1. Diet composition was determined using standard procedures (AOAC, 1999). The amount of feed offered (dry matter basis) was 10% body weight in the fi rst week and 5% body weight from the second week onwards.
Daily feed allowance was offered twice a day (09:00h and 16:00h). For the 50% duckweed + 50% feed diet, duckweed was offered in the morning and the commercial feed in the afternoon.

Evaluation of fi sh performance
Fingerlings in each experimental unit were weighed and measured three times during the experimental period for estimation of average weight gain (fi nal weightinitial weight), feed conversion ratio (total dry feed fed /total wet weight gain) and specifi c growth rate [(100 x Ln fi nal average weight -Ln initial average weight)/ days].

Water quality monitoring
Water temperature and dissolved oxygen concentration were recorded twice a day. Other parameters, such as pH, turbidity, alkalinity, and total ammonia concentration, were determined weekly. The average water pH remained around 6.0; alkalinity around 20mgL -1 total ammonia concentration averaged 0.5mgL -1 during the experimental period. Water transparency determined by Secchi disc varied between 20 and 40cm. Water dissolved oxygen and temperatures recorded throughout the 50-day experimental period are shown in fi gure. 1. The average dissolved oxygen in the morning was lower than in the afternoon and declined as tilapia grew, but not to a concentration that was harmful for tilapia production. However, water temperature averaged 24.9°C in the fi rst period, falling to 23.1°C in the second period, and fi nally dropping to 18.3°C in the third period.

Statistical analyses
Data were analyzed by one way ANOVA, and Tukey test was used to compare differences among treatment means. The level of signifi cance adopted was 5%.

Results and Discussion
Final average weight Final average weight was not different between fi ngerlings fed commercial feed (21.6g) and 50% feed + 50% dried duckweed (19.6g) (p> 0.05) (Figure 1). However, both diets promoted a higher fi nal weight of the fi ngerlings (p<0.05) when compared to dried duckweed alone (fi nal weight 14.4g). A study conducted by Mohedano et al. (2005) reported no differences in weight gain between Nile tilapia fi ngerlings fed only on commercial feed and those fed on formulated feed (of similar nutritional composition) containing 30% duckweed meal, for 60 days. Figure 2 shows that the weight gain was not different between fi ngerlings fed on commercial diets and those fed on 50% feed + 50% dried duckweed (3.70g and 3.53g, respectively) in the fi rst growth period from March 22 nd to April 5 th . However, in the following two periods, fi ngerlings fed on the commercial feed presented higher weight gain (7.79g and 7.76g, respectively). During the three feeding periods, fi ngerlings fed on 50% feed + 50% dried duckweed presented a higher weight gain than the ones fed on dried duckweed alone.

Weight gain and feed conversion
Water temperature decreased throughout the trial from 24.9°C to 18.3°C (Figure 3). Therefore, average weight gain data were plotted in three feeding periods (Figure 2).
Lower water temperature caused a decrease in diet consumption during the second feeding period (April 5 th to April 22 nd ), especially for fi sh fed on dried duckweed alone. However, in the last growth period (April 22 nd to May 13 th ), a signifi cant amount of duckweed was not consumed due to the even lower water temperature. As a result of decreased feed consumption, weight gain decreased for fi ngerlings fed on all diets.
In the fi rst and third feeding periods, the fi ngerlings that received only commercial feed presented the bestfeed conversion, being 1.3 and 2.0 respectively (p<0.05) ( Figure 2). However, in the second period, the average feed conversion ratios for fi ngerlings fed on commercial feed and 50% feed + 50% duckweed were not different (1.7) (p> 0.05). The feed conversion of fi ngerlings fed on dried duckweed alone was always worse than that of fi ngerlings fed on commercial feed or a combination of 50% feed + 50% duckweed. In the fi rst two feeding periods, the feed conversion ratio of fi ngerlings fed on dried duckweed alone ranged from 1.8 to 2.0, while for those fed on 50% feed + 50% dried duckweed, it varied from 1.4 to 1.7.
According to Hasan and Edwards (1992), the feed conversion ratio for tilapia fed on fresh duckweed ranged from 1.6 to 3.3, while tilapia fed on commercial feed in a re-circulation system showed a feed conversion  (Granoth and Porath, 1983). Therefore, during the fi rst two feeding periods, the feed conversion ratios obtained in the present study are within the range reported by other authors. The worst feed conversion and lowest gain presented by fi ngerlings fed on duckweed alone in the third period (Figures 1 and 2) could be explained by the low water temperature recorded at that time. In May, the average water temperature was 17.2°C in the morning and 19.6°C in the afternoon. According to Lim (1989), tilapia activity and feeding becomes reduced when the water temperature is below 20ºC, and feeding stops around 16ºC. In the present study, fi ngerlings fed on commercial feed showed a better performance than fi sh fed on dried duckweed during the third feeding period only probably because the feed pellets sank and it was necessary for the fi sh to come up to the surface in order to eat.

Specifi c growth rate
During the whole feeding trial, the specifi c growth rate of fi ngerlings fed on commercial diet (3.72%) was superior to that presented by fi sh fed on 50% feed + 50% duckweed or duckweed alone (3.02% and 3.30%) (p<0.05). The specifi c growth rates found in the present study are relatively higher than those found in other studies with tilapia. Gaigher et al. (1984) obtained a specifi c growth rate ranging from 1.19 to 1.84% for tilapia fed on fresh duckweed and commercial feed. The reason for this difference could be the high water content of fresh duckweed, which might reduce diet intake and consequently reduce growth. Additionally, the present experiment was conducted outdoors and the fi ngerlings had access to natural feed present in the tank, which may have contributed to the higher growth rates observed.

Duckweed nutritional content
The crude protein and fi ber contents of the dried duckweed used in the present study (38.86% and 13.22%, respectively) are similar to duckweed composition data determined by Gijzen and Khondker (1997). According to these authors, the protein content ranged from 30 to 40%, and the fi ber content from 5 to 15%, when the duckweed was cultivated in nutrient rich media. Swine manure was used to fertilize and cultivate the duckweed in the present study, providing satisfactory growth when it was added at a proportion of 1kg/1000L/week. Once dried, duckweed can be stored in plastic bags, and no other treatment or process is necessary to use it directly as feed for tilapia. However, it is desirable to keep it stored in a cool environment (4°C) before using it for feeding. Furthermore, dried duckweed presents adequate stability in water, fl oating for up to 3h. Duckweed protein production is higher than that observed for most vegetable crops if an annual productivity of 17.6ton dry matter/ha/year and 37% protein concentration is assumed (Iqbal, 1999). Duckweed productivity can vary from 10 to 30ton dry matter/ha/year. These fl uctuations are due to species, climatic conditions, management, available surface area and available nutrients. Mgabwu and Adeniji (1988) reported that duckweed production in nutrientrich water bodies, such as sewage or animal waste lagoons, can reach up to 14.9 g dry matter/m 2 /day if proper management is adopted. Also, the production of duckweed in wastewater can be an alternative which contributes to the sustainability of sewage treatment plants, mitigating environmental impacts. According to Decamp and Warren (2000), duckweed can be used as a secondary treatment of effl uents from activated sludge systems or stabilization ponds since it contributes with nutrient uptake from the wastewater, improving the effl uent quality.
It has already been demonstrated that duckweed can serve as an alternative dietary source for tilapia (El Sayed, 1999;Fasakin et al., 1999;Mohedano et al., 2005). However, most growth studies evaluating fresh duckweed as a dietary item have shown low weigh gain results (Gaigher et al., 1984;Hasan and Edwards, Dried duckweed as an alternative dietary source for tilapia 1992). Gaigher et al. (1984) evaluated the use of fresh duckweeds as feed for tilapia in a water recirculating unit and observed that fi sh fed only on duckweed presented lower weight gain. According to these authors, fresh duckweed intake was limited due to its high moisture content (96%) and the presence of air pockets in its leaves, which are used for fl otation. The present study, however, is the fi rst to evaluate dried duckweed fed directly to tilapia. Our fi ndings show that up to 50% of a 40%-crude-protein commercial feed could be replaced by dried duckweed without reducing the growth of tilapia fi ngerlings for a period of at least 50 days. In the present study, the fi nal average weight of tilapia (from fry to juvenile) fed on commercial pellets was not different when the diet included 50% dried duckweed fertilized with swine manure. However, the low water temperature recorded in May was limiting for growth, negatively affecting the weight gain of fi ngerlings fed dried duckweed alone. Therefore, it is not advisable to rely on dried duckweed as the only food source when water temperatures are below 18°C. Considering that feed can represent up to 70% of total production costs in tilapia culture in semi-intensive systems, feeding tilapia on a 1:1 mixture of dried duckweed and commercial feed could reduce production costs. However, the practical feasibility of such a feeding regime depends on duckweed production costs, which can vary depending on the resources available on each farm.