Cultivation of Gracilaria (Rhodophyta) in shrimp pond effluents in Brazil

Introduction

Intensive aquaculture has contributed to environment degradation, with visible effects such as increases in particulate organic matter and chemical changes such as dissolved oxygen reduction and increased nitrogen and phosphorus concentrations in water (Beveridge 1996). In Brazil, shrimp culture has expanded considerably in the last few years, mainly in the north-east region, which accounts for 97% of total production. In 1999, 15 000 ton of shrimp were produced from 5000 ha of ponds, with predictions of 105 000 ton from 35 000 ha of cultivated area in 2003 (Câmara 2000). The main species of marine shrimp cultivated commercially in Brazil is the white shrimp, Lithopenaeus vannamei, from the Pacific American coast.

In general, shrimp culture produces a large amount of waste, including dissolved nitrogen and phosphorus, that is released to the aquatic environment without treatment. To explore the use of these elements as sources of nutrients, and at the same time to reduce discharges to the environment, seaweed cultivation in shrimp pond effluents appears to be a viable approach. Indeed, seaweeds have been identified as marine plants capable of treating animal culture effluents efficiently (Qian, Wu, Wu & Xie 1996; Troell, Ronnback, Halling, Kautsky & Buschmann 1999; Nelson, Gleen, Conn, Moore, Walsh & Akutagawa 2001; Jones, Dennison, & Preston 2001).

Polyculture of several species of animals and seaweeds has been used traditionally in aquaculture. Such systems are based on the concept that excretions of one organism provide the nutrients for another (Buschmann, Troell, Kautsky & Kautsky 1996). Previous studies have showed enhanced growth of seaweed and shrimp in coculture (Chiang 1981; Shan & Wang 1985; Wei 1990). In combined cultures, productivity depends on the growth performance of all species in the system. Thus, to obtain high productivity, environmental conditions must be favourable for both species (Qian et al. 1996). When conditions are not adequate for one of the species, integrated culture can produce negative results.

Although in Brazil the development of shrimp culture has been rapid, until now there has not been commercial cultivation of seaweeds. While several experiments on the cultivation of Gracilaria had been conducted (Câmara-Neto 1987; Oliveira 1990), cultivation for commercial purposes has not been carried out. At present, only some red algae (Gracilaria and Hypnea) are harvested from natural beds for commercial and subsistence use. However, overexploitation has resulted in depletion of natural beds and consequent shortage in the supply of raw material.

In the present study, the red seaweed Gracilaria sp. was selected because of its availability and economic potential. This species is considered an important source of agar and is harvested from natural populations in Rio Grande do Norte littoral areas. This study is the first evaluation of productivity and growth rate of Gracilaria sp. in shrimp farm effluents.

Material and methods

The experiment was conducted from December 2000 to April 2001 in shrimp pond effluents in Rio Grande do Norte state (06°18′S−35°09′W), Brazil. The shrimp farm has 40 ha of ponds made of clay and sand, with a depth ranging between 1.0 and 1.5 m. Water input into ponds was accomplished using pumps while the water output to effluents was achieved by gravity. The mean stocking density was 25 shrimp per m², with a mean productivity of 4000 kg ha⁻¹ year⁻¹. Feeding rates (10–3%) based on a percentage of total biomass declined with an increase in mean individual shrimp weight. During shrimp culture, a series of daily procedures were conducted, including monitoring of physical–chemical water parameters (temperature, salinity, pH, dissolved oxygen and turbidity). This study was restricted to a drainage canal (effluents), which was 1.5 m deep and 8.0 m wide. The ponds were fertilized with urea, monoammonium phosphate and calcium nitrate. The first fertilization occurred during the filling of pond. Additional fertilizations were applied according to water color and Secchi disc visibility.

Gracilaria was cultivated on six modules, each comprising a square frame of 1.0 × 1.0 m made of PVC pipes and polypropylene rope (5 mm). The modules were placed into effluent. One series of three modules was positioned in the middle of the effluent, while the other other modules were placed near the discharge. They were suspended horizontally at 0.3 m below the water surface with buoys at each extremity, and anchored to the bottom by ropes tied to concrete blocks. The ropes were tied across the frame and fronds of Gracilaria were inserted between the ropes.

Biomass was collected fortnightly, cleaned, weighed and replaced into modules after readjustment to starting biomass (810 g). Restocking was done using cultivated plants, and new thalli from benthic populations were used only when frond loss occurred. Because of the high proportion of small clay particles in the water, modules were gently stirred every 2 days to avoid silt accumulation over fronds. The formula used for determination of relative growth rate (RGR) is presented below (De Casabianca, Marinho-Soriano & Laugier 1997). Although sometimes RGR is not related to specific factors, it is used to express growth in general terms (McLachlan & Bird 1986). Daily growth rate was calculated by the formula:

where w_f is the final weight, w_i is the initial weight and t_f minus t_i is the time interval in days between the two weight measurements.

Water temperature, salinity, dissolved oxygen, water transparency and pH were monitored daily by farm personnel as part of a monitoring routine program. The nutrients (NH₄, NO₃ and PO₄) as well as the biomass were sampled (in triplicate) fortnightly and analysed according to Strickland and Parsons (1972).

To look for significant differences in biomass, RGR and environmental parameter values over the studied period, one-way analysis of variance (anova) was conducted, and differences among biomass, RGR and environmental parameter mean values were tested using Duncan's new multiple range test. Data taken fortnightly were utilized for anova, with n = 30. Levene's test was conducted to verify data variability, and this test showed that the variance was constant (P > 0.05) so that anova could be performed. Pearson correlation coefficient was calculated to determine the linear relationship between variables. To investigate significant differences between the values of the cultivation initial inoculum and the biomass growth after 15 days, the Mann–Whitney U non-parametric test was carried out. This non-parametric test was utilized because the number of observations between those 15 days was small, resulting in problems regarding anova assumptions. anova was performed using the SAS (Statistical Analysis Software), and the Mann–Whitney U-test and Pearson correlations were calculated using the Statistica software.

Results

All environmental parameters measured during this study are summarized in Table 1. During the study period, water temperature in the effluent ranged from 28 °C to 32 °C, and salinity varied from 30‰ to 43‰, with corresponding means of 29.6 ± 1.22 °C and 35.7 ± 3.7‰ for the study period. Salinity showed a highly significant variation (P < 0.01). The concentration of dissolved oxygen fluctuated between 4.1 mg L⁻¹ and 6.8 mg L⁻¹, and water transparency showed a mean of 0.37 ± 0.04 m. pH varied significantly during the study period, with a mean of 7.9 ± 0.31.

Fluctuation of nutrients values was influenced mainly by pond fertilization frequency. Ammonia (NH₄) was the most abundant nutrient and varied from 0.4 mgL⁻¹ to 20 mgL⁻¹; nitrate (NO₃) fluctuated between 2.97 mg L⁻¹ and 8.0 mgL⁻¹, while phosphate (PO₄) oscillated between 0.1 mgL⁻¹ and 0.8 mg L⁻¹. The biomass production between months varied significantly (P < 0.01), reaching a maximum of 2540 g m⁻² in December and a minimum of 380 g m⁻² in April. The mean for the study period was 1419 ± 708 g m⁻². The higher production values always occurred during the first 15 days of cultivation and, on several occasions, exceeded the initial inoculum by 190%. This result, tested by Mann–Whitney U-test (non-parametric test), demonstrated that there were significant differences between initial biomass and the increase in the first 15 days (P < 0.05). The production loss in subsequent days was partially related to module handling. Analysis of variance (anova) showed that there were significant changes in RGR values in the studied months (F_cal = 127.09; P < 0.05). This result was confirmed by Duncan's test. The growth rate of Gracilaria sp. in effluents varied from 8.8% day⁻¹ (December) to 1.8% day⁻¹ (April). During the second April fortnight all the cultivated Gracilaria grew poorly or died. This episode coincided with the use of calcium hypochlorite for pond cleaning. Correlations between RGR and the environmental parameters (water temperature, salinity, pH, water transparency, dissolved oxygen and nutrients) during the study were not significant (P > 0.05).

Discussion

The seaweed production mean obtained in the effluent was 1418 g m⁻² (wet weight). Thus, approximately 28.4 ton of fresh seaweed can be cultivated (per month) per hectare, equivalent to 3.97 ton of dry weight. Annual production in this case would reach 23.93 t ha⁻¹ year⁻¹ (dry weight).

The highest production value reached in our experimental cultivation always occurred during the first 15 days and on several occasions exceeded the weight of the initial inoculum by 2.5 times. After this period there was a clear tendency towards biomass reduction. This biomass reduction can be related to the low water movement and large amount of suspended particles in effluents. These factors contributed to the deposition of a thin layer of silt over thalli, which probably blocked the incident light and affected growth. Similar observations have been reported by other authors (Hurtado-Ponce, Samonte, Luhan & Guanzon 1992; Nelson et al. 2001). Frequent cleaning and maintenance of modules could have partially contributed to thalli fragmentation and consequent biomass loss.

Although the production of biomass of 380–2540 g m⁻² was relatively low compared with tank culture (Ugarte & Santelices 1992), it was superior to values found in ponds by Shang (1976) in Taiwan and Hurtado-Ponce et al. (1992) in the Philippines. The different productivity values observed are mainly as a result of differences between systems and cultivated species. For example, Phang, Shaharuddin, Noraishah & Sasekumar (1996) showed that the growth of Gracilaria changii in an irrigation canal of a shrimp farm was three times higher compared with production in shrimp ponds and/or the natural habitat.

In general and under different environmental conditions, the growth rate from most Gracilaria species is in the range of 5–10% day⁻¹ (McLachlan & Bird 1986). However, during extended periods of growth, rates decreased. The high growth rate (8.8% day⁻¹) obtained in the present study was similar to the rate suggested by McLachlan and Bird (1986) for experiments over short periods of time. Indeed, our observations corroborate their results, considering that highest rates were recorded in the first 15 days (more than 100% of productivity). The growth rate of Gracilaria sp. was similar to that obtained in other experiments (Lignell & Pedersén 1987; Rueness, Mathisen & Tanager 1987; Anderson, Levitt & Share 1996; Kaladharan, Vijayakumaran & Chennubhotla 1996; De Casabianca et al. 1997) but inferior to rates found in effluents by Nelson et al. (2001).

Environmental factors are known to influence the growth of macroalgae (Santelices & Doty 1989; Bird 1988; Capo, Jaramillo, Boyd, Lapointe & Serafy 1999). In the present study, most of these parameters were relatively constant during the studied period and were not correlated to biomass or RGR, indicating that factors other than those reported could have affected biomass production. The subsequent death of most seaweeds in April (tallus deterioration and disintegration) is difficult to explain, considering that there were few changes in environmental parameters. However, this event occurred after the cleaning of one of the ponds using calcium hypochlorite. Indeed, various chemicals used in aquaculture are toxic not only to animals, but also to plants (Primavera 1993).

Aquatic animal culture provides a large amount of nutrients to the environment in the form of metabolite waste (Beveridge 1996). However, it was observed that, despite the increase in the nutrient concentrations during the last few months, productivity did not reach exceptional levels. This was probably due to the excessive increase of ammonia concentration observed in the last months of culture. Indeed, there is evidence from previous studies that seaweed growth can be inhibited by high ammonia concentrations (Waite & Mitchell 1972; Parker 1982; Lignell & Pedersén 1987). Lapointe and Ryther (1979) recorded a decrease in growth rate when seaweeds were exposed to high nitrogen levels, probably due to ammonia toxicity.

The occurrence of higher growth rates in the first days of cultivation is associated with the fact that the seaweeds were collected in non-eutrophicated areas, and probably with low internal contents of total nitrogen and phosphorus. According to the literature, the nutritional state of seaweeds influences the kinetic absorption of nutrients (DeBoer 1981; Fujita 1985; Lobban, Harrison & Duncan 1985), so seaweeds with low nitrogen and phosphorus contents rapidly absorb those nutrients, and consequently exhibited fast growth during the first few days.

Finally, according to the results of this study, we conclude that Gracilaria sp. cultivation can be developed with relative success in shrimp farm effluents. However, we should point out that cultivation success is related to the techniques that are used and to the environmental conditions of the effluent. Ideally, different techniques should be tested and that the culture of organisms (shrimp/seaweeds) should be initiated at the same time, so decreasing the risk of seaweed degradation by harmful substances used in pond cleaning. Further studies in large scale and over longer time periods are necessary to confirm the feasibility of seaweed cultivation at the commercial level in shrimp farm effluents.

Acknowledgments

This work was supported by a research grant from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Brazil. The authors thank the two anonymous reviewers for their comments and corrections.