SciELO - Scientific Electronic Library Online

vol.46 issue2Antioxidant and anti-cancer activities of brown and red seaweed extracts from Chilean coastsStock assessment of jumbo squid Dosidicus gigas in northwest Mexico author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links

  • On index processCited by Google
  • Have no similar articlesSimilars in SciELO
  • On index processSimilars in Google


Latin american journal of aquatic research

On-line version ISSN 0718-560X

Lat. Am. J. Aquat. Res. vol.46 no.2 Valparaíso  2018 

Research Article

Small-scale fish farming in seasonal ponds in Brazil: technical and economic characterization

Adriana Ferreira Lima1 

Ana Paula Oeda Rodrigues1 

Patrícia Oliveira-Maciel1 

Adriano Prysthon1 

Roberto Manolio Valladão-Flores1 

Tácito Araújo-Bezerra2 

1Embrapa Pesca e Aquicultura, Empresa Brasileira de Pesquisa Agropecuária-EMBRAPA, Palmas, Tocantins, Brazil

2Instituto de Desenvolvimento Rural do Tocantins, Palmas, Tocantins, Brazil


The inclusion of aquaculture in household agriculture has been related as an opportunity of product and income diversification as wells as to increase food security for farmers. Small-scale fish production has increased in Brazil, even in regions with a pronounced dry season. However, there is no information about the characteristics and technical and economic viability of aquaculture under these conditions. This study was carried out with the objective to characterize the production of tambatinga (♀Colossoma macropomum × ♂Piaractus brachypomus) in water deficit conditions in Tocantins State, Brazil. The study revealed that it is technically possible and economically feasible to produce tambatinga in seasonal ponds. Worth noting was the adaptability of the hybrids which showed a suitable growth and tolerance to the poor water quality conditions. However, in order to achieve production success and financial returns, the adoption of the following procedures are suggested: a) fish stocking as soon as the rainy period starts, b) use of larger juveniles to achieve a higher final weight, c) adoption of pond liming and fertilizing practices previous to stocking, d) water quality monitoring, e) evaluate fish growing performance, f) partial harvesting during the production cycle, and g) total harvesting at the end of the rainy season.

Keywords: Colossoma macropomum; Piaractus brachypomus; household agriculture; deficit hydric; tambatinga; production


Family agriculture predominates in rural farms in Brazil (IBGE, 2007). In this type of production, farmers are responsible for production management, emphasize production diversification, and use family labor, which may be supplemented with hired labor (MDS, 2014). Production diversification was originally adopted to meet the subsistence needs of the family and is currently used as a strategy to reduce risk and uncertainty (Buainain et al., 2003) by generating income throughout the year, depending on the different crops and productions adopted (Lima, 2008).

In this sense, fish farming stands out as a possible alternative for inclusion in family agriculture due to its high productivity and importance in fighting food insecurity by increasing the consumption of fish, which has high nutritional value, and by supplementing the family income, decreasing poverty, and promoting social and economic development (Ahmed & Lorica, 2002; Baccarin et al., 2009; Kawarazuka & Béné, 2010; FAO, 2011). In addition, it contributes to the better use of the farms’ natural resources and may increase family welfare if used for leisure (Kubitza & Ono, 2010).

Fish farming has been adopted in family farms in several regions of Brazil (Tinoco, 2006; Baccarin et al., 2009; Cardoso et al., 2009). Many of these farms are located in regions with long periods of water deficit, which directly affect aquaculture production (Tucci et al., 2001; INMET, 2014). Due to the seasonality of water availability, farmers seek production alternatives such as shorter fish production cycles (Gupta, 2001; Roos et al., 2002). However, although fish farming is strongly affected by water availability, it shows promise even in locations with water restrictions. This is the case of in the west-central region of the state of Tocantins, Brazil, where family fish farming has been increasing greatly over the last few years as a food and income alternative, promoted by the favorable climate (high temperature and rainfall during the rainy season) and the potential for developing aquaculture in this state (Silva et al., 2013; Torati et al., 2014). Roundfish is the second most produced fish group in Brazil (Rabobank, 2013; IBGE, 2014). The group is represented by the species Piaractus brachypomus (Cuvier, 1818) (pirapitinga), P. mesopotamicus Holmberg, 1887, Colossoma macropomum (Cuvier, 1816) (tambaqui), and their respective hybrids and is the most common fish group produced in seasonal ponds in this region. In this group, the tambatinga hybrid (♀ C. macropomum x ♂ P. brachypomus) is notable. Although little studied, this hybrid is widely produced in the north and west-central regions of Brazil because it combines desirable characteristics from the two parental species such as the fast growth rate of the tambaqui and the high deposition of the dorsal muscle of the pirapitinga (Hashimoto et al., 2012). Tambatinga reaches between 1.5 and 2 kg in one year in regions without a water deficit, but its production potential in seasonal ponds is not known.

To identify technological and socioeconomic improvements and formulate incentive policies to fish farming, it is important to characterize fish production systems (Nhan et al., 2006; FAO, 2011). Using a participatory approach, production systems can be characterized and analyzed together with the farmers involved, integrating their social, cultural, and economic reality (Townsley, 1996; Silva et al., 2013). The present study monitored a production cycle of tambatinga in family farms located in the west-central region of Tocantins, with the aim of characterizing fish production in regions with long periods of water deficit and evaluating the economic contribution of fish production to farmers.


Study site and approach

The west-central region of the state of Tocantins -the municipalities of Divinópolis and Abreulândia (Fig. 1)- was chosen for study due to its history of inclusion of fish farming in family agriculture and long periods of water deficit (about six months), where fish farming characteristically occurs in seasonal ponds (Gupta, 2001).

Figure 1 Location of the study site. 

The study was based on a preliminary participatory appraisal performed to understand the dynamics of local family fish farming (Silva et al., 2013). Visits were conducted before the study began to raise awareness and engage the local farmers in helping with the monitoring. The entire study was conducted using a participatory approach, where the local farmers actively participated in the selection of the fish farms to be monitored and in the fish production, definition of the production parameters, and data collection and recording.

Selection of fish farms to be monitored

A workshop was held with local farmers during which the aims of the study and the specific activities that would have to be performed were explained. Following the workshop, the farmers themselves indicated seven farms to be monitored during one fish production cycle, taking into consideration the interest of each farmer in participating in the study and the availability of the farmers to perform the necessary activities. In a second workshop directed only at the selected farmers, the farmers were trained on how to complete the forms, use instruments, and monitor production using the necessary techniques, according to the learning-by-doing principle (De et al., 2012). This was followed by individual training at the farms, as described by Lima et al. (2014).

Production monitoring

To allow comparisons among the different production units, the following technical production aspects were standardized with the help of the farmers: a) date of the beginning of monitoring, based on the time from when the seasonal ponds had sufficient water to begin fish farming, b) fish species, to facilitate comparisons regarding fish performance and economic aspects, and c) stocking density, considering the region's history of total mortality due to high stocking densities (which would make monitoring impossible), water availability throughout the year, production characteristics of the species, and fish consumption by the family during the production cycle.

Seven family farms, with nine seasonal ponds, were monitored (Table 1). Production began on 12 December 2012, using tambatinga juveniles weighing 1.49 ± 0.60 g. Fish fry were obtained from a fish farm located in the study region in the municipality of Brejinho de Nazaré, transported in plastic containers, and acclimatized to the ponds. The initial reference stocking density was 1 fish m-2 and was adjusted to a higher density (1.22 ± 0.33 fish m-2) when the producer expressed interest in consuming fish during the production cycle (farms A, B, D, and E) or lower when they expressed a lack of financial ability to maintain a large number of fish (farms C and F) (Table 1).

Table 1 Seasonal ponds monitored during one production cycle of tambatinga (♀ Colossoma macropomum x ♂ Piaractus brachypomus). Pond surface area, stocking density, water quality parameters (mean ± standard deviation), and fish mortality for the different studied production units. 

Production unit* Pond surface-area (m2) Stocking density (fish m2) Temperature (°C) Water transparency (cm) Dissolved oxygen (mg L–1)*** pH Total ammonia (mg L−1)* CO2 (mg L-l) Alkalinity (mgL−1) Nitrite (mgL−1) Mortality (%) Period of data collection
Initial Final
A1 192 1.56 27.6 ±1.2 91.2 ± 37.5 1.2 ±0.9 6.4 ±0.3 0.5 ± 0.2 17.9 ±6.3 21.3 ±3.6 0.1 ±0.3 0.33 12/16/2012 08/31/2013
A2 288 1.46 27.4 ± 1.3 50.6 ± 15.4 2.6 ± 1.2 6.8 ±0.4 0.6 ±0.0 9.8 ±4.4 25.1 ±8.0 0.0 ±0.0 10.95 12/17/2012 08/31/2013
B1 250 1.10 27.4 ± 0.7 63.6 ± 19.0 3.4 ±0.6 6.0 ± 0.2 0.6 ± 0.2 14.4 ±2.4 9.7 ±4.9 0.2 ± 0.1 0.36 12/15/2012 06/16/2013
B2 270 1.03 27.4 ± 0.8 66.4 ± 12.1 3.9 ±0.5 6.0 ±0.3 0.5 ± 0.2 14.8 ±3.0 10.4 ±5.6 0.2 ±0.1 47.67 12/15/2012 06/16/2013
C 800 0.63 28.5 ± 1.0 51.4± 17.2 2.8 ± 1.0 7.0 ± 0.0 0.6 ± 0.0 6.1 ±0.5 8.4 ± 2.0 0.0 ± 0.0 3.35 02/07/2013 06/02/2013
D 236 1.53 26.5 ± 2.6 33.4 ±6.2 NQ 6.9 ±0.6 1.5 ±2.3 20 ± 9.3 50.5 ±22.0 0.3 ± 0.2 0 12/15/2012 07/08/2013
E 200 1.00 28.3 ± 1.1 57.2 ± 18.6 1.8 ±0.5 6.7 ±0.5 0.6 ± 0.2 16.4 ±3.5 31.2 ±7.3 0.0 ± 0.0 2.5 12/15/2012 08/10/2013
F 11.555 0.14 28.9 ± 1.1 48.2 ± 8.6 NQ 6.5 ±0.5 0.6 ± 0.0 9.7 ±0.7 21.3 ± 1.2 0.0 ± 0.0 7.58 01/14/2013 08/10/2013
G 125 1.44 31.5 ± 1.6 NQ NQ 8.0 ± 1.0 0.3 ± 0.3 2.5 ±3.0 43.5 ± 24.0 0.3 ± 0.2 1.11 12/16/2012 03/10/201.3

NQ: not quantified;

*Different letters indicate different farms, and different numbers indicate more than one production unit within a given farm.

**Collection of the dissolved oxygen data began on 16 May 2013.

***All the production units presented toxic ammonia below 0.02 mg L−1.

During the production cycle, the fish were fed commercial feed for omnivorous fish. Only two farmers (C and F) reported using other ingredients (ground soybean, corn, and bean), and these were used for less than 15 days. The following technical and economic data were collected in partnership with the farmers:

  1. Fish farming expenses: the use of inputs, purchased and not purchased (acquired from the farm);

  2. Pond liming and fertilization before stocking and during the production cycle: type and amount of input applied in each procedure;

  3. Water quality parameters, measured daily: temperature (in the morning) was measured using a thermometer, transparency (between 11:00 h AM and 12:00 h PM) using a Secchi disk, and pH (in the morning) using a water analysis kit. Dissolved oxygen was initially measured using a water analysis kit, and five months after the beginning of monitoring, in six production units, using digital oximeters;

  4. Water quality parameters, measured weekly in the morning: total ammonia, carbon dioxide, total alkalinity, and nitrite, using a water analysis kit;

  5. Daily feeding: quantity and type of feed or food offered;

  6. Rainfall: only its occurrence or absence was recorded;

  7. Decrease in the number of fish at the fish farm: due to mortality, consumption, or sale; and

  8. Monthly biometry measurements: fish weight (in groups), standard length, and health.

The monitored farms were visited monthly for reviewing, discussing, and monitoring the data collected by the farmers. The biometric measurements were performed during these visits, with the help of the research team. In addition, with the aim of stimulating the exchange of experiences between local farmers, participating or not in the study, monthly workshops were held to discuss fish farming, including its difficulties and solutions.

Economic analysis

Because the farmers on some fish farms harvested the fish when the fish began to lose weight, an economic analysis was performed, taking into consideration two time periods: 1) the actual period, which corresponded to the total production cycle, and 2) the ideal period, which corresponded to the ideal harvest time, i.e., the time when the fish was at the highest weight in each fish farm. For fish farms where the harvest was performed when the fish were at their highest weight, the ideal period was not estimated.

The production cost was calculated considering the components described by Matsunaga et al. (1976): the effective operational cost (EOC): the expenses of fish fry, feed, correctives, fertilizers, and maintenance of machinery and equipment; the total operational cost (TOC): the effective operational cost plus expenses with the depreciation of improvements; and the total cost (TC): the total operational cost plus the opportunity cost of labor.

The indicators of economic viability of production used were gross income (GI): income from the production sales; net margin (NM): the ratio between the net income and total operational cost; and profit (P): the difference between the gross income and total production costs, including all production factors (Scorvo-Filho et al., 2004).

Labor was calculated as the opportunity cost because services were performed by the farmers themselves, and no outside labor was hired. Labor was therefore considered as that of a multifunctional worker, working for one hour per day during the fish production cycle, plus a total of 16 h for fish stocking and harvest. The value of work hours was calculated based on the current minimum wage.

Statistical analysis

All statistical analyses were performed using the R software (R Core Team, 2013). The different production units were divided into three groups: I) with liming and fertilization, II) only with fertilization, and III) without pond preparation procedures (Karim et al., 2011). Differences in fish growth during the initial growth phase between the three groups of production units were tested using analysis of variance, considering the categorized production units as the independent variable, followed by Tukey's test, at P < 0.05. Relations among the different water quality parameters measured, time of cultivation and fish weight were analyzed using Pearson's correlation coefficient. Multiple stepwise linear regression was used to identify which variables best explained the variance in fish weight.


Fish farm characteristics

Most fish farms studied were set in ponds originally built as water sources for livestock, serving as reservoirs for direct or indirect (resulting from water table rising) water catchment, without water entrance or exit, with depth varying between 1.5 and 4.0 m and a mean area of 223 m2 (except on farms C and F) (Table 1).

Time of cultivation

Cultivation began approximately 30 to 45 days after the first rains in the region when the seasonal ponds contained sufficient water to begin fish farming. Water availability became limited in most farms at the beginning of the dry season. The end of the production cycle varied for different production units, extending to between 170 and 270 days of cultivation, depending on the water dynamics of each production unit. High soil drainage in production unit G made fish production impracticable, and the production cycle was terminated during the third month of cultivation.

Water quality and rainfall

The mean values for the water quality parameters, measured daily or weekly, are presented in Table 1. Production unit D had high mean ammonia concentration, and production unit G had high pH. Water transparency could not be measured in production unit G due to high soil drainage, low fertility, and a low water level. Units B1, B2, and C had low alkalinity.

The variations in water quality parameters along the production cycle are presented in Fig. 2. Temperature and pH were the only relatively stable parameters throughout the production cycle, exhibiting very little variation compared to the remaining parameters. Water transparency was initially high in most production units and gradually decreased along the production cycle.

Figure 2 Time series of the dissolved oxygen, carbon dioxide, water transparency, pH, temperature, and nitrite for tambatinga (♀ Colossoma macropomum x ♂ Piaractus brachypomus) production in seasonal ponds. 

Significant linear correlations were observed between some water quality parameters, although with low correlation coefficients (Table 2). The pH was significantly correlated with all water quality parameters except ammonia. Likewise, dissolved oxygen was significantly correlated with all water quality parameters except ammonia and alkalinity. In contrast, ammonia was only significantly correlated with nitrite and alkalinity.

Table 2 Pearson correlation coefficients for the relationships between water quality parameters during one production cycle of tambatinga (♀ Colossoma macropomum x ♂ Piaractus brachypomus) in seasonal ponds. 

Water transparency 0.04
Dissolved oxygen 0.29* 0.29*
pH 0.48*** -0.25* 0.26*
Ammonia -0.17 -0.01 0.10 -0.01
Carbon dioxide -0.21 0.47*** -0.28* -0 47*** -0.08
Alkalinity 0.19 0.46*** 0.15 0.46*** 0.45*** -0.06
Nitrite 0.14 0.07 0.34* 0.07*** -0.25* 0.08 0.08

T: temperature; WT: water transparency; DO: dissolved oxygen; A: ammonia; CD: carbon dioxide; Alk: alkalinity; N: nitrite.

*P < 0.05;

**P < 0.01;

***P < 0.001.

The occurrence of rainfall in the study region was high during the first four months of cultivation, becoming less frequent during the fifth month, and ceasing between the sixth and seventh month. Rainfall occurred between four and seven months, determining the duration of the production cycle at the different farms (Table 3).

Table 3 Rainfall frequency during one production cycle of tambatinga (♀ Colossoma macropomum x ♂ Piaractus brachypomus) in seasonal ponds. 

Month Monthly rainfall frequency (%) per farm
December NQ 58.82 NQ NQ NQ NQ
January 52.38 61.29 NQ 42.86 25.00 NQ
February 32.14 59.26 42.86 28.57 35.71 47.83
March 41.94 51.61 54.84 35.48 51.61 29.03
April 23.33 33.33 23.33 13.33 26.67 26.67
May 6.45 19.35 6.45 6.45 9.68 6.45
June 0.00 6.67 6.67 0.00 3.33 0.00
July 0.00 0.00 * * 0.00 0.00
August 0.00 * * * 0.00 0.00
September 0.00 * * * * *
End of production cycle/monitoring** September June(B2**)and July (B1**) June June August August

NQ: not quantified.

*Month following the end of the production cycle or monitoring.

**Different letters indicate different farms, and different numbers indicate more than one production unit within a given farm.

Fish performance and mortality

Fish performance varied significantly among fishponds. During the initial grow-out phase (53 days of cultivation), in general, the increase in fish weight was highest for the production units that received liming and fertilization, followed by those that received only fertilization, and was lowest for production units that received no preparation procedures (P < 0.01) (Table 4). The procedures of liming and fertilization were performed prior to fish stocking. A significant correlation between fish weight following 53 days of cultivation and water quality parameters was only found for pH (0.876; P < 0.01).

Table 4 Mean initial (at 53 days of cultivation), maximum and final fish weight, daily weight gain, time of cultivation, and fish pond preparation procedures performed. 

Production unit* Pond preparation Initial weight (g) Weight at 53 days(g) Maximum weight (g) Ideal period** (days) Final weight (g) Actual period (days) Daily weight gain (g day−1)***
Liming Fertilization
A1 Yes Yes 1.5 96.3a 760 208 549 271 3.65
A2 Yes Yes 1.5 66.7a 519 208 328 271 2.50
B1 Yes Yes 1.5 64.8a 214 174 214 174 1.23
B2 No Yes 1.5 59.3ab 152 153 135 174 0.99
C No Yes 1.5 57.14ab 261 153 170 174 1.70
D No Yes 1.5 46.5ab 465 173 465 173 2.69
E No Yes 1.5 43.4ab 463 208 449 244 2.23
F No No 1.5 42.7b 401 173 401 173 2.32
G**** No No 1.5 27.5b 172 90 172 90 1.91

*Different letters in the same column indicate different farms, and different numbers indicate more than one production unit within a given farm.

**Time between the beginning of the production cycle and the time when the animals reached the highest weight.

***Calculated for the period between the beginning of the production cycle and the time when the animals reached the highest weight.

****Production cycle ended early due to problems with the production unit.

abMeans followed by different letters within the same column are significantly different according to Tukey's test (P < 0.05).

The weight variation and weight gain per month of the tambatinga during its production cycle in the monitored fishponds are presented in Figure 3, and the maximum and final weight and daily weight gain are presented in Table 4. Fish grew until approximately 150 to 200 days of cultivation (between the fifth and the sixth month), when they reached their maximum weight of 152 to 760 g.

Figure 3 Weight growth curve (line) and weight gain (bars) of tambatinga (♀ Colossoma macropomum x ♂ Piaractus brachypomus) grown in seasonal ponds. The “x” indicates the last rainfall. A, B, C, D, E, F, G indicate the different farms, and different numbers indicate more than one production unit within a given farm. *Cultivation ended early due to problems with the production unit. 

This was followed by a period of weight loss, beginning after the last rainfall event in some farms, resulting in final weights between 135 and 549 g for a cultivation time between 90 and 271 days (Table 4).

The mean tambatinga daily weight gain until the animals reached their highest weight was 2.13 g. Except for farm D, where the fish were transferred to another pond located in the same farm that does not dry out, fish weight increased until 120 days of cultivation and then decreased, coinciding with the decrease in rainfall observed starting in the fifth month. Between 170 and 200 days of cultivation, weight gain became negative, indicating fish weight loss in most production units. In farm E, weight gain was again positive in July (seventh month), following the removal of 47 fish in June to decrease the stocking density. However, fish weight decreased again during the eighth month of cultivation.

The analysis of the correlations between fish weight variation along the production cycle and water quality parameters and time of cultivation revealed that the fish weight variation along the production cycle was significantly correlated (P < 0.05) with the time of cultivation (0.8412), dissolved oxygen (0.6977), carbon dioxide (0.5249), and nitrite (0.3057). However, the correlation coefficients were low for the correlations with carbon dioxide and nitrite (Table 5).

Table 5 Correlation between weight and water parameters of tambatinga (♀ Colossoma macropomum x ♂Piaractus brachypomus) production in seasonal ponds. 

Dependent variable Independent variable Pearson's correlation coefficient P-value
Weight Temperature -0.2380902 0.07722
Period (month) 0.8412025 2.22e-16***
Water transparency 0.2580406 0.05958
Dissolved oxygen -0.6977079 6.404e-09***
pH -0.1939196 0.1521
Ammonia -0.06622796 0.6375
Carbon dioxide 0.524862 5.468e-05***
Alkalinity -0.04297589 0.7599
Nitrite -0.3057114 0.02601*

*P < 0.05;

**P < 0.01;

***P < 0.001.

Multiple stepwise linear regression analysis, with fish weight as the dependent variable and the time of cultivation and water quality parameters as the independent variables, revealed that the fish weight variation was best explained by a model containing five variables (Table 6).

Table 6 Multiple regression model for weight and independent variables. 

Dependent variable Independent variable B P-value
Weight Intercept -346.45 0.062
Period (month) 47.17 1.76 e-05***
Water transparency 1.13 0.066
Dissolved oxygen -10.21 0.028*
pH 39.86 0.067
Carbon dioxide 21.660 0.002**

R2 = 0.7831; P-value = 6.299e-14.

*P < 0.05;

**P < 0.01;

***P < 0.001.

The accumulated fish mortality during the production cycle remained under 10% for most production units (Table 1). The fish mortality observed in production unit B2 was related to fish health problems caused by the protozoan ectoparasite Piscinoodinium pillulare (Schäperclaus, 1954) in the sixth month of cultivation. The mortality event lasted four days, with a total loss of 131 fish (48.5%). Necropsies were performed, and P. pillulare was identified according to Martins et al. (2015). No treatment was applied because mortality decreased (43, 39, 25, and 24) and completely ceased on the fourth day. Overall, mortality was concentrated in the first month of cultivation in all production units.

Fish consumption

Fish consumption by farmers and their families occurred in all farms monitored in the present study, corresponding to between 2 and 48% of the total production (Fig. 4). Regarding commercialization, all farmers were able to sell part or all their production to the local community, even if the production cycle was terminated when the fish was still very small.

Figure 4 Amount of fish produced in the studied farms and fish consumption by the farmers. 

Economic evaluation of the production

Considering the actual production period, fish farming resulted in losses for most farmers (Table 7). However, eight of the nine production units presented a positive net margin, showing that fish farming was economically unviable if the opportunity cost of labor was taken into account. B2 was the only farm that presented a negative net margin, which was due to the high mortality rate caused by the disease observed in the sixth cultivation month. When the ideal production period was considered (the period during which the production units exhibited their maximum production potential), five of the nine production units could have obtained profit from fish farming, even considering the farmers’ labor as the opportunity cost of labor. In addition, the highest profit was obtained in farm F, which stocked the highest number of fish and therefore diluted the opportunity cost of labor.

Table 7 Overall summary of the cost of production of tambatinga (♀ Colossoma macropomum x ♂ Piaractus brachypomus) in seasonal ponds for two different harvest times. Actual period - total production cycle; Ideal period - estimated for a hypothetical scenario with harvest performed when the fish reached the highest weight. 

Production unit* Production cost (US$)
Actual period Ideal period
A1 388.08 209.03 473.01 179.05 -84.93 536.26 198.68 412.19 337.58 124.07
A2 288.61 200.09 463.10 88.51 -174.43 458.95 193.72 406.26 265.23 52.69
B1 138.88 87.91 270.37 50.97 -131.49
B2 42.37 82.32 264.77 -39.95 -222.40
C 193.77 141.95 272.96 51.82 -79.19 297.15 141.95 252.58 155.20 44.57
D 395.04 131.42 312.91 263.62 82.14
E 206.76 141.14 390.55 65.62 -183.80 213.01 122.42 336.90 90.60 -123.88
F 1379.72 545.10 724.65 834.62 655.08
G** 72.04 42.13 141.12 29.91 -69.08

Values are in US Dollars (US$ 1 = R$ 3.39)

GI: gross income, TOC: total operational cost, TC: total production cost (including the opportunity cost of labor), NM: Net margin, the difference between GI and TOC, P: profit considering the difference between GI and TC.

*Different letters indicate different farms, and different numbers indicate more than one production unit within a given farm.

**Production was ended early due to problems in pond structure.

The feed was the most expensive cost item in fish farming, representing 75 to 91% of the effective operational costs. Fish fry was the second most expensive item, representing between 6 and 22% of the production costs (Table 8).

Table 8 Percentage participation of each cost item on the effective operational cost of production of tambatinga (♀ Colossoma macropomum x ♂ Piaractus brachypomus) in seasonal ponds. 

Production unit* Participation in the effective operational cost (%)
Feed Fish fry Correctives Fertilizers Maintenance of machinery and equipment
A1 85.99 6.64 6.19 0.70 0.49
A2 81.95 10.06 6.71 0.75 0.53
B1 77.24 15.11 7.32 0.33
B2 75.37 16.41 7.87 0.35
C 82.09 16.40 0.77 0.55 0.20
D 86.29 12.81 0.69 0.21
E 91.00 8.80 0.20
F 82.49 17.33 0.18
G 75.68 22.73 0.84 0.76

*Different letters indicate different farms, and different numbers indicate more than one production unit within a given farm.

The feed was the most expensive cost item in fish farming, representing 75 to 91% of the effective operational costs. The fish fry was the second most expensive item, representing between 6 and 22% of the production costs (Table 8).


Fish farm characteristics

The mean area of fish farms in the study region is within the recommended area for family fish farming (300 m2) because it allows the production of enough fish to meet the family demands and of surplus for commercialization and has lower construction costs (Gopalakrishnan & Coche, 1994; Carballo et al., 2008). Larger cultivation structures can stock larger numbers of fish and, consequently, have higher production costs, which in many cases would not be suited to the financial capacity of the families. However, larger structures would enable higher production volumes and dilute the opportunity cost of labor, as discussed in section “Economic evaluation of the production”.

The use of livestock watering ponds for fish farming is common in many regions in Brazil (Garutti, 2003), allowing farmers to optimize the use of their production infrastructure. Considering that small bodies of water can have multiple uses in rural properties (Little et al., 2006) and that production diversification is a characteristic of family agriculture (Buainain et al., 2003), the use of livestock watering ponds for multiple purposes on the farms reinforces the production diversification characteristic of family farms. However, this integration of the production space would be more efficient if fish farming was planned before the watering ponds were constructed, enabling the inclusion of the necessary technical requirements for fish farming, such as water entrance and exit and appropriate depth and slope. The high depth of the ponds studied results from the need to maintain the water in ponds for longer during the dry season, as previously reported for India by Kumar (1992). However, the high depth made the fish harvest for monthly biometric measurements and consumption by the farmer more difficult.

Although these technical deficiencies in pond construction do not prevent the development of fish farming in the studied region, they hinder fish production management and performance. For example, the construction of ponds for fish farming in soil with high drainage, together with the farmer's low resources and lack of input alternatives in the region for waterproofing the pond bottom, led farmer G to end production early in the third month. Fishpond E was built so that it received water from a dam reservoir with 6,000 m2, thereby maintaining 1.5 m water depth during the dry season, but this water supply ceased during the dry season due to the insufficient slope. In the fish pond, C, the impossibility of completely draining the pond for its disinfection before fish stocking resulted in a high number of invasive fish (mainly Geophagus sp.), which competed with tambatinga for food and oxygen, have no commercial value and are not traditionally consumed by farmers as food. Therefore, planning of the production units so that they can be used for both fish farming and other activities is important.

Water quality and rainfall

Temperature and pH were stable during the production cycle and within the recommended values for fish production (Boyd & Lichtkoppler, 1979; Baldisseroto & Gomes, 2010). Only production unit G presented a higher mean pH than the remaining production units, but this pH was still within the recommended range. Because liming was performed when the ponds were still empty of water and before stocking, the water alkalinity level for each unit was not known. Production unit G, therefore, would not have required liming, which resulted in increased water pH and alkalinity. This was also the case with production units D, E, and F, which presented alkalinity within the indicated range for fish production even without liming. Two of three production units that received liming also presented alkalinity within the indicated range for fish farming.

Water transparency varied among the production units, initially due to the fertilization performed. In spite of variations, the overall water transparency remained within the recommended values for fish farming, being initially higher and decreasing along the production cycle. This change in transparency was expected for fish farming and probably explains the inclusion of water transparency in the regression model obtained for the fish weight variation (Table 6). Total ammonia and nitrite remained at adequate values for fish production even at the end of the production cycle when there was no more water renewal, and the overall concentrations of nitrogen compounds were higher (Boyd & Tucker, 1998; Faria et al., 2013; Moro et al., 2013). The observed ammonia concentrations indicated that without water limitation, fish production could have been maintained for a longer period, even without water renewal.

Some production units presented high water carbon dioxide concentrations but within the recommended values for fish farming (Boyd & Tucker, 1998). The dissolved oxygen concentration was one of the limiting factors for fish performance, as it was lower than recommended during almost the entire production cycle, especially during the final months (Silva et al., 2007; Gomes et al., 2010; Faria et al., 2013; Moro et al., 2013). It reached critical levels several times, resulting in changes in fish behavior, with the fish rising to the surface in search of better oxygen concentrations at the water-air interface. This behavior is an adaptive mechanism called aquatic surface respiration (ASR) and is associated with the expansion of the lower lip, which funnels the thin surface layer of oxygen-rich water toward the gills (Val & AlmeidaVal, 1995). However, no mortality was observed from water-quality problems, showing the resistance of this hybrid, similar to that of the parental tambaqui, to low dissolved oxygen concentrations (Silva et al., 2007; Gomes et al., 2010).

The correlations observed among the water quality parameters are characteristic of the water quality dynamics in fish farming environments (Boyd & Lichkoppler, 1979). The increase in nutrient input to the fishpond environment due to feeding supply results in increased phytoplankton populations through the production cycle and, consequently, in higher production of oxygen and carbon dioxide. This explains the correlations observed among water transparency and pH, carbon dioxide, and dissolved oxygen. Temperature, in turn, affects photosynthesis and respiration and, consequently, the dissolved oxygen level and pH. The ammonia concentration was not correlated with pH or dissolved oxygen, possibly due to the low variation in the ammonia concentrations during the production cycle; ammonia concentrations were therefore only correlated with nitrite and alkalinity. The correlation observed between pH and alkalinity is explained by the fact that alkalinity is an expression of the buffering capacity of water, and the correlation between pH and nitrite concentration is explained by ammonia nitrification, which decreases the pH. Ammonia nitrification also explained the significant correlation between ammonia and nitrite concentrations.

In fish farming in seasonal ponds, the production cycles are naturally shorter. Fish species that can reach adequate weight for commercialization during these shorter periods should be selected. The production cycle for roundfish is usually one year (Gomes et al., 2010), and fish farming in seasonal ponds at the study region was only possible for six to seven months. However, this shorter period was sufficient for tambatinga to reach the necessary size for commercialization at the local community. Similar production cycle durations have been reported for family fish farming in seasonal ponds in Bangladesh, showing that shorter production cycles are not an impediment to fish farming in seasonal ponds in Brazil (Gupta, 2001; Roos et al., 2002).

Fish performance

Liming and fertilization were associated with better animal performance during the first 53 days of cultivation, resulting in a significant correlation between weight and pH for this period. The liming of ponds for fish farming is performed with the aim of neutralizing pH, increasing alkalinity, and promoting biological productivity, thereby improving the water quality for fish production (Kungvankij & Chua, 1986). The aim of fertilization is to improve the natural fertility of ponds to allow the development of zooplankton (Karim et al., 2011; Mischke, 2012), which are naturally consumed by tambaqui (Goulding & Carvalho, 1982; Silva et al., 2003). The fish farms that received both procedures, therefore, exhibited better fish performance, reaching a mean fish weight of 67.61 g at the end of 53 days. This weight was higher than reported by Silva et al. (2007) for tambaqui grown for 60 days in fishponds that had received liming and fertilization, which reached a final weight of 46.47 ± 1.4 g. This reinforces the importance of these procedures to improve animal performance during this stage of cultivation and to optimize fish growth within limited production cycles due to water restrictions. Karim et al. (2011) studied the production of Nile tilapia (Oreochromis niloticus, Linnaeus, 1758) in Asia and concluded that to increase productivity in family fish farming, increasing the nutrient concentrations in fish ponds is more viable than increasing the stocking densities, which confirms the importance of this type of management.

The failure to adopt practices of pond preparation for fish farming is, in most cases, related to a lack of knowledge about fish farming (Karim et al., 2011). The access to the necessary inputs for these procedures is not usually a limiting factor, as some of these inputs can be acquired or are already routinely used in family farms for breeding of other animals and small-scale agriculture.

The fish weight loss following the last rainfall in the region resulted from worsening of the water conditions (especially dissolved oxygen), which resulted from the impossibility of renewing the pond water. In addition, after the last rainfall, the water volume at the ponds began to decrease, resulting in higher number of fish per water volume, a higher stocking density, and critical oxygen levels, which may have also contributed to fish weight loss.

Regardless of the weight-gain variations among the different production units, the observed overall weight gain was higher than that reported by Izel & Melo (2004) (1.69 g d−1), and Paula (2009) (1.25 g d−1), showing the positive response of the animals to the studied production system. However, the overall weight gain was lower than that reported by Padilla (2000) (1.1-1.8 g d−1), and Arbeláez-Rojas et al. (2002) (4.5 g d−1), indicating that the management practices adopted can be improved to maximize the production potential of the species, even under water-deficit conditions.

Overall, the highest fish weights, between 401 and 760 g, were reached between 170 and 200 days of cultivation, when the last rains occurred in the region. This would be the ideal time for harvest. Until that time, the overall fish growth was within the expected values for the species (SEBRAE, 2013). However, most fish were not harvested at this time and lost weight, exhibiting lower weight at the end of the production cycle than expected for tambaqui grown in ponds without water deficit, which can reach from 1 kg in 270 days (Oliveira et al., 2004) to 2 kg in 360 days (SEBRAE, 2013) from an initial weight of 1 g.

In India, De et al. (2012) observed that beginning cultivation with larger fish fry is a possible strategy to optimize the rainfall period in regions with water deficit or in ponds with low water retention capacity, resulting in higher final biomass. This could also be a possible strategy to optimize production in the studied production systems.

The strong correlation between fish weight and cultivation time was expected. The correlation between fish weight and dissolved oxygen, carbon dioxide, and nitrite concentrations reflected the importance of these parameters to the fish development. Dissolved oxygen concentrations outside the fish comfort zone decrease fish growth rates (Boyd & Lichtkoppler, 1979). The level of dissolved oxygen is usually more critical at the end of the production cycle due to the high fish biomass, which explains its correlation with fish weight. Similarly, carbon dioxide production increases with increasing fish biomass due to increased fish respiration rate.


The fish mortality observed in the studied production units was within acceptable levels for fish production (Izel & Melo, 2004; EMBRAPA, 2006; SEBRAE, 2013). Higher mortality during the first month of cultivation was expected, as this is the stage when fish are more sensitive and when the highest mortalities naturally occur (Chagas et al., 2005; Gomes et al., 2010).

High mortality was only observed in one production unit, and this was caused by parasitosis. The protozoan P. pillulare parasitizes fish gills and skin, causing velvet disease. The parasite's reproduction is favored by the poor water quality caused by high fish densities and excess organic matter (Martins et al., 2015). Martins et al. (2001) reported that hybrids such as tambacu suffer higher rates of parasitism by P. pillulare than pure species such as Piaractus mesopotamicus and Colossoma macropomum. In the present study, the velvet disease outbreak was associated with decreased or cessation of food consumption by fish seven days before the first day of mortality. Hyporexia is a classic indicator of health or water quality problems (Noga, 1996). In addition, biometric measurements were performed two days before the outbreak, which was an additional source of stress to the fish and may have contributed to decreased fish immunity, making the fish susceptible to the presence of the parasite in the environment.

Although the water quality was poor at the end of the production cycle, this did not cause tambatinga mortality in most production units, as tambatinga are quite resistant and adapted to regions with water deficit.

Fish consumption and commercialization

Fish consumption by farmers along the production cycle varied widely among the different production units but occurred for all units, confirming the importance of including fish farming in family agricultural systems as a way of introducing high-quality protein into the diets of the families. This is one of the most important contributions of family fish farming (Muir & Nugent, 1995).

Buainain et al. (2003) also observed variations in the degree of consumption between farms depending on the production system adopted and the level of capitalization of farmers. The authors observed that even among the farmers with the highest level of capitalization in southern Brazil, family consumption corresponded to almost 20% of the total fish farm production. In some areas of northern and northeastern Brazil, this percentage is considerably higher, as in some production units in this study, reflecting not only the precariousness of the means available to farmers but especially their isolation and distance from markets. In the studied region, fish farming is not the predominant production activity, and the fish that are traditionally consumed originate from fishing and, consequently, depending on seasonality. However, the inclusion of fish farming in the studied farms allowed fish consumption throughout the production cycle, even when the fish were still smaller than the commercial size.

Both fish consumption and fish commercialization at the local community occurred when the animals exhibited weight lower than the common market weight. Farmer C, for example, was able to sell his entire production when the fish had a mean weight of 170 g, showing the existence of a local market for fish under 200 g. This should be taken into account for this production system when planning harvests and sales along the production cycle because it allows farmers to gradually decrease fish density and generate income. The farmers who participated in the present study were able to sell part or all of their production to the local community, therefore contributing to the local food security.

Economic evaluation of production

Underwater deficit conditions, fish farming may be a profitable activity for family producers if correctly planned and managed. At the studied farms, the losses resulting from fish farming resulted from inadequate management, where the fish were harvested when exhibiting productivity loss. However, even in these cases, if the opportunity cost of labor was not taken into account, the fish production presented a positive net margin. If the farmers had performed the harvest at the ideal time, five of the nine production units would have obtained profit from the activity, even accounting for the opportunity cost of labor. This is a positive outcome for family producers, as, in addition to improving the family diet, fish farming may generate additional income and diversify income sources, increasing the economic security of farmers.

Regarding the production costs of a fish farm, the feed is the most expensive item (Martin et al., 1995; Scorvo-Filho et al., 2004; Vera-Calderón & Ferreira, 2004; Andrade et al., 2005). This was also observed in the present study. For Pangasianodon hypophthalmus (Sauvage, 1878) production in intensive, semiintensive, and extensive systems, feed costs represented a high percentage of the total production costs but decreased with decreasing system intensity (75.8, 68.8, and 58.5%, respectively) (Ahmed et al., 2010). In spite of the benefits of using industrial feeds, the cost of feed is still quite high for small producers (Tacon et al., 2011; Ahmed & Toufique, 2014). In the studied region, however, the use of industrial feed is the most appropriate to fully exploit the growth potential of these fish, as fish performance and the production cycle are limited by water restrictions. Adopting systems that are more extensive would result in longer production cycles, which are not compatible with the local water availability. The cost of fish fry was the second-costliest item. The fish fry is usually an important item for production costs and deserves special attention from producers. Although the cost of fertilizers is low, this cost may be further decreased if the producer uses inputs obtained at the farm, such as organic fertilizers (Nhan et al., 2006; Karim et al., 2011).

Considering the economic evaluation of the studied farms, the importance of organizing the producers in associations for the purchase of inputs at lower prices should be highlighted. In addition, a positive factor for the studied group of fish farmers is the use of family labor because, in some cases, fish farming resulted in positive economic outcomes due to the lack of wage payment.

Karim et al. (2011) attributed differences in production and economic performance among small farmers producing tilapia under polyculture systems to differences in education, knowledge, and experience. These differences were also observed in the present study, together with the great scarcity of specialized technical assistance in the region. Technical assistance and rural extension programs may help increase farmer knowledge about fish farming and decrease these differences in the future. This is known to be one of the main bottlenecks for the development of fish farming in Brazil (Poli et al., 2000; Ostrensky et al., 2008; CNA, 2010). Increased knowledge about fish farming, together with the adoption of adequate production practices and production periods, is therefore expected to increase fish-farming efficiency and economic return to the families.

Evaluation of the participatory method

Participatory methods have been used in local communities and family farms to facilitate the development of research in collaboration with producers, making it more democratic (Nhan et al., 2006; Karim et al., 2011; De et al., 2012). The training of farmers for the monitoring, which extended from farmer selection by the community to the training workshops, technical meetings, and monthly visits, was of great importance to the success of monitoring and exchange of experiences between farmers and researchers. In addition, the monthly workshops with the farmers participating in the study and with other local farmers resulted in increased social relations, culminating in the establishment of a local farmers association at the end of the study.


This study was funded by the Brazilian Agricultural Research Corporation (EMBRAPA). The authors are thankful to Rosiana Rodrigues Alves for the collaboration with the statistical analysis and producers who participated in the present research.


Ahmed, M. & M.H. Lorica. 2002. Improving developing country food security through aquaculture development lessons from Asia. Food Policy, 27: 125-141. [ Links ]

Ahmed, N. & K.A. Toufique. 2014. Greening the blue revolution of small-scale freshwater aquaculture in Mymensingh, Bangladesh. Aquacult. Res., 46: 2305-2322. [ Links ]

Ahmed, N., M.F. Alam & M. Hasan. 2010. The economics of sutchi catfish (Pangasianodon hypophthalmus) aquaculture under three different farming systems in rural Bangladesh. Aquacult. Res., 41: 1668-1682. [ Links ]

Andrade, R.L.B., R.L. Wagner, I. Mahl & R.S. Martins. 2005. Custos de produção de tilápias (Oreochromis niloticus) em um modelo de propriedade da região oeste do Estado do Paraná, Brasil. Cienc. Rural, 35: 198-203. [ Links ]

Arbeláez-Rojas, G.A., D.M. Fracalossi & J.D.I. Fim. 2002. Composição corporal de tambaqui, Colossoma macropomum, e matrinxã, Brycon cephalus, em sistemas de cultivo intensivo, em igarapé, e semi-intensivo, em viveiros. Rev. Bras. Zootec., 31: 1059-1069. [ Links ]

Baldisseroto, B. & L. de C. Gomes. 2010. Espécies nativas para piscicultura no Brasil. UFSM, Santa Maria, 608 pp. [ Links ]

Baccarin, A.E., A.F.G. Leonardo, L. Tachibana & C.F. Correia. 2009. Piscicultura em comunidade remanescente de quilombo: um estudo de caso. Inf. Econ., 39: 42-47. [ Links ]

Boyd, C.E. & F. Lichtkoppler. 1979. Water quality management in pond fish culture. Auburn, Res. Dev. Ser., 22: 30 pp. [ Links ]

Boyd, C.E. & C.S. Tucker. 1998. Aquaculture water quality management. Kluwer Academic Publishers, Boston, 700 pp. [ Links ]

Buainain, A.M., A.R. Romero & C. Guanziroli. 2003. Agricultura familiar e o novo mundo rural. Sociologias, 10: 312-347. [ Links ]

Cardoso, E.S., H.M.O. Rocha & M.C. Furlan. 2009. A piscicultura no município de Santa Maria, R.S. Cienc. Nat., 31: 131-140. [ Links ]

Carballo, E., A.V. Eer, T.V. Schie & A. Hilbrands. 2008. Small-scale freshwater fish farming. Agromisa Foundation an CTA, Wageningen, 84 pp. [ Links ]

Chagas, E.C., L.C. Gomes, H. Martins Júnior, R. Roubach & J.N.P. Lourenço. 2005. Desempenho de tambaqui cultivado em tanques-rede, em lago de várzea, sob diferentes taxas de alimentação. Pesqu. Agropec. Bras., 40: 833-835. [ Links ]

Confederação Nacional da Agricultura (CNA). 2010. Diagnóstico nacional da atividade aquícola, Confederação Nacional da Agricultura e Pecuária do Brasil. CNA, Brasília, 12 pp. [ Links ]

De, H.K., D.N. Chattopadhyay, R. Radheyshyam, G.S. Saha, A.K. Dash, S. Pal & T.S. Satpati. 2012. Strengthening the livelihoods of rural women through polyculture of carps in seasonal ponds. Indian J. Fish., 59: 137-141. [ Links ]

Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA). 2006. Cultivo de Peixes. EMBRAPA Informação Tecnológica, Brasília, 25 pp. [ Links ]

Food and Agriculture Organization of the United Nations (FAO). 2011. World aquaculture 2010. FAO, Rome, 120 pp. [ Links ]

Faria, R.H.S.A., M. Morais, M.R.G.S. Soranna & W.B. Sallum. 2013. Manual de criação de peixes em viveiros. Codevasf, Brasília, 136 pp. [ Links ]

Garutti, V. 2003. Piscicultura ecológica. UNESP, São Paulo, 336 pp. [ Links ]

Goulding, M. & M.L. Carvalho. 1982. Life history and management of the tambaqui (Colossoma macropomum, Characidae): an important Amazonian food fish. Rev. Bras. Zool., 1: 107-133. [ Links ]

Gomes, L.C., L.N. Simões & C.A.R.M. Araújo-Lima. 2010. Tambaqui (Colossoma macropomun). In: B. Baldisseroto & L.C. Gomes (eds.). Espécies nativas para piscicultura no Brasil. UFMS, Santa Maria, pp. 175-204. [ Links ]

Gopalakrishnan, V. & A.G. Coche. 1994. Handbook on small-scale freshwater fish farming. FAO, Rome, 215 pp. [ Links ]

Gupta, M.V. 2001. Culture of short-cycle species in seasonal ponds and ditches in Bangladesh. In: Integrated agriculture-aquaculture. FAO, Rome, pp. 45. [ Links ]

Hashimoto, D.T., J.A. Senhorini, F. Forest & F. Porto-Foresti. 2012. Interspecific fish hybrids in Brazil: management of genetic resources for sustainable use. Rev. Aquacult., 4: 108-118. [ Links ]

Instituto Brasileiro de Geografia (IBGE). 2007. Censo Agropecuário 2006, resultados preliminares. IBGE, Rio de Janeiro, 141 pp. [ Links ]

Instituto Brasileiro de Geografia (IBGE). 2014. Produção da pecuária municipal 2013. IBGE, Rio de Janeiro, 108 pp. [ Links ]

Instituto Nacional de Meteorologia (INMET). 2014. Dados da Rede do INMET. []. Reviewed: 16 March 2017. [ Links ]

Izel, A.C.U. & L.A.S. Melo. 2004. Criação de tambaqui (Colossoma macropomum) em tanques escavados no Estado do Amazonas. Embrapa Amazônia Ocidental, Manaus, 24 pp. [ Links ]

Karim, M., D.C. Little, M.S. Kabir, M.J.C. Verdegem, T. Telfer & M.A. Wahab. 2011. Enhancing benefits from polycultures including tilapia (Oreochromis niloticus) within integrated pond-dike systems: a participatory trial with households of varying socio-economic level in rural and peri-urban areas of Bangladesh. Aquaculture, 314: 225-235. [ Links ]

Kawarazuka, N. & C. Béné. 2010. Linking small-scale fisheries and aquaculture to household nutritional security: an overview. Food Security, 2: 343-357. [ Links ]

Kubitza, F. & E.A. Ono. 2010. Piscicultura familiar como ferramenta para o desenvolvimento e segurança alimentar no meio rural. Panorama Aquicult., 20: 1423. [ Links ]

Kumar, D. 1992. Fishculture in undrainable ponds. A manual for extension. FAO, Rome, 239 pp. [ Links ]

Kungvankij, P. & T.E. Chua. 1986. Shrimp culture: pond design, operation and management. NACA Training Manual 2, FAO, Phillippines. []. Reviewed: 15 April 2017. [ Links ]

Lima, S.O.S. 2008. Sistemas de produção e processos de trabalho no campo: caderno pedagógico educandas e educandos. Ministério da Educação, Brasília, 159 pp. [ Links ]

Lima, A.F., A.P. Silva, A.P.O. Rodrigues, M.X. Pedroza-Filho, P.O. Maciel, R.M.V. Flores & T.A. Bezerra. 2014. Metodologia para o monitoramento de dados técnicos e econômicos em pisciculturas familiares. Bol. Pesqui. Desenvol., 1: 172 pp. [ Links ]

Little, D.C., M. Karim, D. Turongruang, E.J. Morales, F.J. Murray, K. Barman, M.M. Haque, N. Kundu, B. Belton, G. Faruque, E.M. Azim, F.U.L. Islam, L.J. Pollock, M.J. Verdegem, J.A. Young, W. Leschen & M.A. Wahab. 2006. Livelihood impacts of ponds in Asia opportunities and constraints. In: A.J. Van der Zijpp, J.A.J. Verreth, M.E.F. Van Mensvoort, R.H. Bosma & M.C.M. Beveridge (eds.). Fishponds in farming systems. Wageningen Academic Publishers, Netherlands, pp. 177-202. [ Links ]

Martin, N.B., J.D. Scorvo-Filho, E.G. Sanches, P.F.C. Novato & L.M.S. Ayrosa. 1995. Custos e retornos na piscicultura em São Paulo. Inf. Econ., 25: 9-47. [ Links ]

Martins, M.L., J.R. Moraes, P.M. Andrade, S.H.C. Schalch & F.R. Moraes. 2001. Piscinoodinium pillulare (Schäperclaus, 1954) Lom, 1981 (Dinoflagellida) infection in cultivated freshwater fish from the northeast region of Sao Paulo State, Brazil. Parasitological and pathological aspects. Braz. J. Biol., 61: 639-644. [ Links ]

Martins, M.L., L. Cardoso, N. Marchiori & S.B. Pádua. 2015. Protozoan infections in farmed fish from Brazil: diagnosis and pathogenesis. Braz. J. Vet. Parasitol., 24: 1-20. [ Links ]

Matsunaga, M., P.F. Bemelmans, P.E.N. Toledo, R.D. Duley, H. Okawa & I.A. Pedroso. 1976. Metodologia do custo de produção utilizada pelo IEA. Agricultura em São Paulo. 23: 123-129. [ Links ]

Ministério de Desenvolvimento Sustentável (MDS). 2014. Agricultura familiar. []. Reviewed: 16 March 2017. [ Links ]

Mischke, C.C. 2012. Aquaculture pond fertilization: impacts of nutrients input on fertilization. Wiley-Blackwell, Mississipi, 308 pp. [ Links ]

Moro, G.V., L.S. Torati, D.B. Luiz & F.T. Matos. 2013. Monitoramento e manejo da qualidade de água em pisciculturas. In: A.P.O. Rodrigues, A.F. Lima, A.L. Alves, D.K. Rosa, L.S. Torati & V.R.V. Santos (eds.). Piscicultura de água doce: multiplicando conhecimentos. Embrapa, Brasília, pp. 141-169. [ Links ]

Muir, J.F. & C.G. Nugent. 1995. Aquaculture production trends: perspectives for food security. FAO Fish. Dept., Rome, 133 pp. [ Links ]

Nhan, D.K., A. Milstein, M.C.J. Verdegem & J.A.V. Verreth. 2006. Food inputs, water quality and nutrient accumulation in integrated pond systems: a multivariate approach. Aquaculture, 261: 160-173. [ Links ]

Noga, E.J. 1996. Fish disease: diagnosis and treatment. Mosby-Year Book, St. Louis, 519 pp. [ Links ]

Oliveira, A.M.B.S., L. Conte & J.E.P. Cyrino. 2004. Produção de Characiformes autóctones. In: J.E.P. Cyrino, E.C. Urbinati, D.M. Fracalossi & N. Castangnolli (eds.). Tópicos especiais em piscicultura de água doce tropical intensiva. TecArt, São Paulo, pp. 217-238. [ Links ]

Ostrensky, A., J.R. Borghetti & D. Soto. 2008. Aquicultura no Brasil: o desafio é crescer. FAO, Rome, 276 pp. [ Links ]

Padilla, P.P. 2000. Efecto del contenido proteico y energético de dietas em el crecimiento de alevinos de gamitana (Colossoma macropomum). Folia Amazon., 10: 81-90. [ Links ]

Paula, F.G. 2009. Desempenho do tambaqui (Colossoma macropomum), da pirapitinga (Piaractus brachypomum), e do híbrido tambatinga (C. macropomum x P. brachypomum) mantidos em viveiros fertilizados na fase de engorda. []. Reviewed: 16 March 2017. [ Links ]

Poli, C.R., A. Grumann & J.B. Borghetti. 2000. Situação atual da aquicultura na região Sul, In: W.C. Valenti, C.R. Poli, J.A. Pereira & J.B. Borghetti (eds.). Aquicultura no Brasil: bases para um desenvolvimento sustentável. CNPq, Ministério da Ciência e Tecnologia, Brasília, pp. 321-351. [ Links ]

R Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL []. [ Links ]

Rabobank. 2013. Brazilian aquaculture. Rabobank Industry Note, 362: 1-11. [ Links ]

Roos, N., S.H. Thisted & M.A. Wahab. 2002. Culture of small indigenous fish species in seasonal ponds in Bangladesh: the potential for production and impact on food and nutrition security. In: D.C. Edwards, D.C Little & H. Demaine (eds.). Rural aquaculture. CAB Internacional, London, pp. 245-252. [ Links ]

Scorvo-Filho, J.D., M.I.E.G. Martins & C.M.D. Scorvo-Frasca. 2004. Instrumentos para análise da competitividade na piscicultura. In: J.E.P. Cyrino, E.C. Urbinati, D.M. Fracalossi & N. Castangnolli (eds.). Tópicos especiais em piscicultura de água doce tropical intensiva. TecArt, São Paulo, pp. 517-533. [ Links ]

Serviço Brasileiro de Apoio às Micro e Pequenas Empresas (SEBRAE). 2013. Manual como iniciar piscicultura com espécies regionais. Sebrae, Brasilia, 80 pp. [ Links ]

Silva, J.A.M., M. Pereira-Filho & M.I. Oliveira-Pereira. 2003. Frutos e sementes consumidos pelo tambaqui, Colossoma macropomum (Cuvier, 1818) incorporados em rações. Digestibilidade e velocidade de trânsito pelo trato gastrointestinal. Rev. Bras. Zootecn., 32: 1815-1824. [ Links ]

Silva, A.M.D., L.C. Gomes & R. Roubach. 2007. Growth, yield, water and effluent quality in ponds with differentmanagement during tambaqui juvenile production. Pesqu. Agropec. Bras., 42: 733-740. [ Links ]

Silva, A.P., A.P.O. Rodrigues, D.N. Sousa, P.O. Maciel, M.X. Pedroza Filho, M.E. Ummus & R.M.V. Flores. 2013. Diagnóstico participativo da piscicultura familiar na região de Divinópolis/TO: Uma abordagem diferenciada para ações de pesquisa e desenvolvimento. Embrapa Pesca e Aquicultura, Palmas, 96 pp. [ Links ]

Tacon, A.G.J., M.R. Hasan & M. Metian. 2011. Demand and supply of feed ingredients for farmed fish and crustaceans: trends and prospects. FAO, Rome, 87 pp. [ Links ]

Tinoco, S.T.J. 2006. Análise socioeconômica da piscicultura em unidades de produção agropecuária familiares da região de tupã, SP. 2006. []. Reviewed: 16 March 2017. [ Links ]

Torati, L.S., A. Pristhon, A.P.O. Rodrigues, A.F. Lima, P.O. Maciel, M.X. Pedroza-Filho & M.E. Ummus. 2014. Uso da água e caracterização ambiental de pisciculturas familiares do noroeste do Tocantins, Brasil. Rev. Iberoam. Cienc. Amb., 5: 96-106. [ Links ]

Townsley, P. 1996. Rapid rural appraisal, participatory rural appraisal, and aquaculture. FAO Fish. Tech. Pap, 358: 109 pp. [ Links ]

Tucci, C.E.M., I. Hespanhol & O.M. Cordeiro-Netto. 2001. Gestão da água no Brasil. UNESCO, Brasília, 178 pp. [ Links ]

Val, A.L. & V.M.F. Almeida-Val. 1995. Fishes of the Amazon and their environment. Physiological and biochemical features. Springer, Heidelberg, 224 pp. [ Links ]

Vera-Calderón, L.E. & A.C.M. Ferreira. 2004. Estudo da economia de escala na piscicultura em tanque-rede, no estado de São Paulo. Inf. Econ., 34: 7-17. [ Links ]

Received: May 10, 2017; Accepted: October 25, 2017

Corresponding author: Adriana Ferrerira Lima (

Corresponding editor: Mauricio Laterça

Creative Commons License This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.