- Email: [email protected]
Transport limitation of oxygen in shrimp culture ponds
Aquacultural Engineering 10 ( 1991 ) 269-279
Transport Limitation of Oxygen in Shrimp Culture Ponds A l b e r t G a r c i a III Agricultural Engineering Department, Texas A&M University, College Station. Texas 77843, USA
D a v i d E. B r u n e Agricultural & Biological Engineering Department, Clemson University,Clemson, South Carolina 29634, USA (Received 9 January 1989; accepted 11 July 1991 ) ABSTRACT A 120-acre shrimp farm located in southeast Texas has failed for three years to produce expected yields of shrimp despite intensive management. Data obtained from these ponds are combined with observations from other culture facilities to support the hypothesis that the degree of mixing of pond water is the primary limitation to oxygenation of the benthic biomass in a shrimp culture pond. The implication of this conclusion is that shrimp ponds are advection limited," therefore, rational design of such facilities will require detailed data concerning the impact of interactions of pond size, depth and geometry with wind speed, and aerator placement, upon pond mixing. INTRODUCTION A set of six, 8 ha (20 acre), ponds approximately 1 m (3-28 ft) deep located on the Texas coast south of Corpus Christi, Texas, were stocked with Penaeus vannamei at different times and stocking densities. Ponds 1 through 4 were stocked on 26 May, 1987 with 0-62 g animals at approximately 105 000/ha ( - 10/mZ). Ponds 5 and 6 were stocked on 30 April, 1987 with 0"8g animals at 4 5 0 0 0 / h a ( - 5 / m 2 ) . Growth proceeded normally in ponds 5 and 6, but was retarded in ponds 1 to 4 (see Table 1 ). 269 Aquacultural Engineering 0144-8609/92/S05.00 - © 1992 Elsevier Science Publishers Ltd, England. Printed in Great Britain
270
A. Garcia Ill, D. E. B r u n e
After 5 weeks, ponds 1 through 4 experienced a significant amount of bird predation which is indicative of excessive mortality. Mortality in these ponds was estimated to be one percent per day. At this time the average shrimp weight in ponds 1-4 was 3"5 g. Ponds 5 and 6 were harvested after 11 weeks. The average weight in these ponds was 13"5 g and the total yield was 540 kg/ha. Oxygen concentration and temperature, which were monitored twice daily, did not indicate periods of extreme stress. The pathology of the animals in ponds 1 through 4 during the growout period did not indicate a singular cause of death. The data showed that the animals contained an increasing amount of opportunistic microflora as time progressed. While it is impossible to ascertain with certainty the cause of mortality in ponds 1 through 4, it is possible to collect facts and hypothesize a likely explanation. The following analysis is presented as an attempt to describe the mechanism for mortality in this particular case, and to propose general research objectives to address the suggested problem. BACKGROUND
Pond oxygen dynamics Oxygen management in aquaculture ponds is made difficult not only by the complex web of ecological interactions but also by diurnal fluctuations. During periods of light, the photosynthetic biomass both produces and consumes oxygen. Usually the production exceeds the light period respiration thus elevating the pond oxygen levels. On calm days, intensive algal bloom often elevates the water oxygen concentration to superTABLE ! Growth (g) Data for Ponds Stocked at 45 000" and 105 000 t' Animals Elapsed time
P o n d 5"
l ' o n d 6"
Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9 Week 10
3.17 4.50 5.23 5-70 6.50 5"90 8.40 9.20 10'9
3.70 5.00 6.23 5.80 6-50 7.80 9-20 9.40 10'0
l ' o n d I t,
t ' o n d 2 I'
Pond 3 h
l'ond 4 h
4-06 5-20 3.50
4-30 4-80 3.40
4"25 4.96 3"40
4'10 4"40 3"80
Oxygen transport in shrimp culture ponds
271
saturated concentrations. During the night, metabolism shifts completely to respiration. The oxygen may be rapidly consumed, with the lowest oxygen levels normally occurring just before dawn. Whenever pond 02 concentrations are below saturation, wind will serve to reaerate the pond with the rate of mass transfer of oxygen a function of wind speed. However, when the pond is at supersaturated conditions the wind will act to de-oxygenate the water. In addition to the diurnal variation in oxygen levels, there is an ever present biochemical oxygen demand (BOD) from added shrimp feed. BOD is exerted in the pond in two zones, one part associated with the water column and a second part associated with the sediment (SOD). The SOD is usually the more significant component affecting the pond water quality in ponds rearing benthic organisms at productivities in excess of 1000 lb/acre (Fast et al, 1988). The sources of the organics driving the SOD are the addition of artificial feeds and fertilizers, influent water and decay of sinking autotrophic biomass. For simplicity of analysis one may consider the sum of animal respiration and the respiration of the fecal matter to be equal to the ultimate oxygen demand of the feed added to the pond. From the perspective of the shrimp, the water near the sediment is the environment in which they must feed and grow, and the water column represents the reservoir of oxygen. In a well mixed and adequately aerated pond, the two are the same and no distinction is necessary. In such cases, advective oxygen mass transfer is sufficient for benthic demand. A typical management practice is to assume the pond is well mixed and to measure the oxygen level at the mid depth in the late afternoon and early morning and assume that such a measurement represents the environment to which the shrimp are exposed.
Oxygen transport limitations A problem may arise when the assumption that the pond is well mixed is no longer valid. During a calm period, advective mass transport may be virtually eliminated or, depending on depth and pond size, advective transfer may be limited to a shallow zone, even during windy days, and may not provide sufficient oxygen transfer, particularly at high densities. At such time, the only mechanism for mass transfer will be diffusion. Under such conditions, the diffusivity of oxygen in water becomes the limiting factor to oxygen transport to the shrimp biomass. A simple way of looking at this is to focus on oxygen flux. Oxygen enters the water at the air-water interface at a rate described by the two-film gas transfer theory (Lewis & Whitman, 1924). It is partially consumed/produced in
272
A. Garcia 11I, D. E. Brune
the water column and diffuses to the sediment. If the SOD, oxygen differential across the diffusive zone, the diffusivity of oxygen in salt water, and the depth of the diffusive zone are known, then the rate of oxygen supply to the sediments can be calculated. Normally it is the reverse calculation that is more revealing, which can be stated as; what is the maximum thickness of an unmixed layer of water that can deliver enough oxygen to the water-sediment interface to meet the shrimp and SOD requirements while maintaining a minimum oxygen level of 0-5 ppm (lethal limit)? To answer this, the SOD must be estimated and a best case may be constructed for the oxygen levels present in the pond. It is known from previous work (Chieng et al., 1988) that the ultimate BOD of the shrimp feed is 650 mg 02/g. At a stocking rate of 45 000 and 105 000 animals/ha, and an overall feeding rate ranging from 1% to 7% of body weight per day, the SOD may be calculated (Table 2). Measured values of SOD, in these ponds, were seen to be in the range of these calculated values (Garcia, 1987). In calculating the 'best case' for diffusive transport of oxygen at the pond sediment interface, the oxygen in the bulk pond water was assumed to be near saturation (8.0 ppm). This would likely represent the oxygen condition on a bright calm day. A level of 0.5 ppm at the sediment-water interface was chosen as the lethal limit to sustain P. vannamei. The thickness of a hypothetical diffusive layer at the sediment water interface may be calculated as follows (see Fig. 1 ):
TABLE 2 Calculated Benthic Oxygen Uptake Rate of Shrimp Ponds (g O2/m -~ day) at Feeding Rates Varying from 7% to 1% of Body Weight per Day Animal weight (g)
1 5 10 15 20 25
Number of animals/m:
Feed rate (%)
5
lO
20
40
lO0
200
0"23 0"82 0"82 0-74 0-98 0-82
0"46 1"64 1'64 1"48 1"96 1'64
0-92 3-28 3'28 2-96 3"92 3"28
1"84 6"56 6'56 5"92 7"84 6"56
4-60 16"40 16"40 14-80 19"60 16'40
9"20 32"80 32"80 29"60 39"20 32"80
7 5 2'5 1"5 1"5 1'0
Oxygen transport in shrimp cultureponds
273
DO m 8.0 m9]l (Saturation)
ADVECTIVE ZONE
~ Z== thloknou
i
/)
.T.
oo = o.s mo/I
DIFFUSIVE
POND S E D I M E N T -
Fig. 1.
ZONE.
,
SOD :
19 m 9 0 l / m 2 / h
C~,,h,Um,O
Representation of diffussive oxygen transport in a shrimp pond.
where F--oxygen flux, mg OE/m 2 day; D=oxygen diffusivity, 2.07 × 10 -4 mE/day; Z = thickness of diffusive zone, m; ( d c / d z ) = c o n centration gradient, mg/m 3. In the case of a pond stocked with 1 g animals at 10 animals/m 2 (an estimated SOD of 460 mg OE/m 2 day) this calculation suggests a diffusive layer limit of 0-34 cm. This is a highly significant statement, indicating the necessity of maintaining mixed conditions in a pond. Even under the most favorable conditions at stocking densities of 105 000 animals/ha (10/mE), a stagnant layer of 0-34 cm would be the limit at which oxygen could diffuse to the sediment rapidly enough to meet the lowest oxygen demand experienced during the 190 day growout period. Therefore, under most conditions, the pond bottom and hence the shrimp are totally dependant upon advective transport for mixing (and 02 transport) to within less than 0.5 cm of the bottom. Response of shrimp to oxygen concentration Previous investigations of penaeid shrimp oxygen response have suggested that the respiratory response of these animals is dependant on a number of factors including oxygen level, animal size, activity and species. For smaller animals (less than 3-0 g) respiration per unit biomass can be more than twice as high than that of larger animals (Egusa, 1961 ). Egusa's data show that large shrimp (greater than 5.0 g) are probably respiratory regulated at oxygen levels of 2.0 rag/liter. Data from Seidman
274
A. Garcia 111, D. E. Brune
and Lawrence (1985) confirm the observation that 2.0 mg/liter is a concentration that will limit growth. It is not unreasonable to assume that concentrations of 2.0 mg/liter may lead to animals stress; this assumption is supported by the observation that raceway culturists prefer to maintain outlet dissolved oxygen levels in excess of 3.5 mg/hter (Wong, 1983). This information suggests that the shrimp respiration in the observed ponds was under constant regulation by the pond oxygen level. Further, if the oxygen flux to the sediment is subject to the degree of mixing of the water column (as suggested by the previous calculation), then it is reasonable to hypothesize that the degree of animal stress (and consequently animal mortality) is also linked to the intensity of pond mixing at increasing carrying capacity. The actual pond culture involves larger shrimp, more sediment oxygen demand, nighttime respiration and lower oxygen levels in the water column. These conditions will likely lead to the imposition of chronic stress, if not, outright catastrophic failure.
DISCUSSION If the degree of pond mixing is limiting shrimp biomass yields from culture ponds, then one would expect that the required horsepower to sustain high density shrimp cultures would be greater than that needed to simply oxygenate the pond. For example, a typical waste treatment lagoon being loaded at a rate of 50 lb (22.7 kg) of BOD/acre day would require approximately 2 hp/acre (0.6 kw/ha) to adequately aerate the lagoon (Loehr, 1977). However, as Loehr observes, such a pond would require nearly 29 hp/acre (8.8 kw/ha) to keep the bacterial and waste solids in suspension. Therefore, there exists an approximately 10-fold difference in power requirements to mix rapidly settling particles as opposed to air-water interfacial oxygen transfer. The logical question to ask is : Does a similar situation exist for shrimp culture ponds; in other words, what is the increased horsepower requirement to maintain a mixed zone at the pond bottom? Table 3 gives the oxygenation rates expected from paddlewheel aerators operating at power levels observed by Hopkins et al. (1989 unpublished) to provide 'adequate' pond oxygen conditions (0.1 hp/animal m2). Such aerators will typically provide between 1.8 and 2.0 lb of O2/hp-h hr (1-2 kg/kw-h) at an oxygen concentration of 0 mg/liter. In Table 3 the rates are adjusted for 3 mg/liter and 5 mg/liter pond 02 levels and 12 and 24 hr aerator operation cycles.
Oxygen transport in shrimp culture ponds
275
TABLE 3 Aerator oxygenation (g O2/m ~ day) at Aerator Power Levels Suggested by Hopkins et al. (1989)
Aerator cycle (h)
Oxygen level (mg/liter)
5
10
20
40
100
2(X)
12
5 3 5 3
0'43 0-68 0"86 1-36
0"86 1-36 1"72 2"72
1"72 2'72 3.44 5'44
3"44 5'44 6"88 10-88
8"60 13"60 17"20 27'20
17'20 27'20 34.40 54.40
0.5
1.0
2.0
4"0
10.0
20.0
24
Recommended aerator power (hp/acre)
Number of animals/m-'
In contrast, wind driven surface re-aeration at a wind speed of 5.4 m/s (average annual wind for the southeast Texas coast) are predicted by equations from Banks and Henera (1977) to be 3.7 gm/m 2 day at 3.0 rag/ liter of pond dissolved oxygen or 2.3 g / m 2 day at 5.0 mg/liter of pond dissolved oxygen levels. Ponds 5 and 6 (see Table 1 ) were unaerated and successfully produced 10 g animals stocked at 5 animals/m 2. At this size and density, Table 2 predicts a maximum SOD of 0.98 g/m 2 day with a wind reaeration potential of 2"3 to 3.7 g / m 2 day or a ratio of (potential/required) aeration of 2"3 to 3.7. In the unsuccessful ponds (1-4) this ratio drops to 1-2-1.8. This suggests that approximately two to four times as much power (supplied by wind) is needed to maintain a 'healthy' pond as would be expected based on aeration requirements alone. In the higher density ponds (20-200 animals/m 2) which are always aerated, the ratio averages 1.4. Again, more power is needed to maintain the pond than is expected (based on aeration requirements) however, the ratio is less, probably because paddlewheel aerators are more effective at mixing the diffusive layer than surface wind action. Figure 2 presents the calculated average shrimp respiration rates in ponds 1-4 and ponds 5 and 6. This figure further supports the oxygen limitation concept. Shrimp respiration represents approximately 10-20% of total feed respiration (calculated from data presented by Chieng et al., (1988) Seidman and Lawrence (1985) and Egusa (1961). Figure 2 graphically demonstrates how the shrimp biomass can continue to expand as long as total pond respiration is at least a factor of 2.0 below pond reaeration (supplied by wind). However, pond 5 and 6 peaked early above this value; the end result being a slow shrimp mortality as the
276
A. Garcia Ill, D. E. Brune
500-
~
400-
PONDS !, 2, 3, 4
>-
300. I
.......
E c~ 200-
E
100-
0 0
Fig. 2.
///
PONDS 5, 6
......... I ......... I ......... t ......... [ 5 10 15 20
TIMES (Weeks) Shrimp oxygen uptake rates in successful (5 and 6) and unsuccessful ( 1,2, 3 and 4) grow-out ponds.
biomass 'self corrects'. If aquatic biomass accumulation is limited by oxygen then an expected response would be for the total mass of shrimp to achieve a maximum limit regardless of stocking density. This phenomena has been demonstrated previously with P. vannamei (Wyban et aL, 1987). This effect can manifest itself in two ways. The first is for growth to continue as normal for part of the population. As the biomass attempts to expand, a certain portion must die in order to prevent the total 02 demand from surpassing the biomass limit. Another manifestation of the same problem is for growth to cease altogether; with the total biomass remaining at maintenance levels. The actual response is usually some combination of the two possibilities. As early as 1970, Broom et al. speculated that the shrimp stocked in large unaerated ponds in excess of 50 000 animals/ha would be limited by the oxygenation capacity of the pond. He also concluded that higher stocking densities and yields would be possible when economical aeration systems were developed. Tables 2 and 3 suggest that this limit is, in fact, 50 to 25% of the true oxygenation capacity, and more closely represents the mixing limit of the pond as a result of wind action. This conclusion is supported by a study conducted by Rogers and Fast (1988) in which they found that freshwater prawn production could be
Oxygen transport in shrimp culture ponds
277
increased 8-11%, at 1600 lb/acre (1800 kg/ha) with induced mixing of less than 0"25 hp acre. Past researchers (Parker et al., 1974) have observed that stocking rates up to 80000 animals/ha (8-0 animal/m 2) in small unaerated, or minimally aerated, 0"1 ha ponds does not produce significant mortality. This suggests that small ponds are better mixed than large ponds. Wind action on small ponds may result in increased mixing as water is pushed against the bank and circulated to deeper layers. In spite of the fact that larger ponds (with a larger fetch) produce greater wind-induced wave height, this apparently does not produce greater bottom water mixing in the pond. It is common practice to position shrimp ponds with short axis into the prevailing wind direction (Losordo, 1989) or at least, to turn the long axis 30 degrees away from the wind (Hanson & Goodwin, 1977). This implies that shore action is necessary for proper mixing; shore length per unit of water volume may be a better indication of pond mixing than pond fetch. The case of 8.0 animal/m 2 represents a maximum SOD of approximately 1.57 g/m 2 day. This suggests a potential/required aeration ratio ranging from 1.46 to 2"36 which suggests that the efficiency of wind mixing in small (0-1 ha) ponds approaches the level of paddlewheel mixing in larger ponds. The efficiency of wind induced mixing in large 8 ha will likely be significantly reduced. In a recent related study, Wyban et al. (1989) demonstrated successful culture of Penaeus vannamei at 25 animal/m 2 in unaerated ponds located at Makapuu Point, Hawaii. The average wind speed at this location (during the growing season) is approximately 16% greater than Corpus Christi, Texas giving an overall wind aeration rate (at 5.0 rag/liter) of 4.48 g / m 2 day as opposed to 3"68 at Texas. At 25 animal/m 2 this gives aeration potential/required ratio of 1"2 which is at the same level which produced failure in the Texas ponds. However, the Texas ponds were 8 ha (20 acre) while the Hawaiian ponds were only 0.4 ha (1 acre). The observation of increased wind-induced oxygenation of Hawaiian ponds is supported by Madenjian et al., (1987). Interestingly, Wyban et al. (1987) found that in the unaerated ponds, individual animal weight had only reached 15 g/animal at a ratio of 1.2 while similar, aerated ponds expanded to 21 g/animal, at which time the ratio again approached 1.25. Due to the very limited data base (as usually is the case) any reasons proposed for the observed Texas shrimp mortality are speculative at best. The authors are well aware of this and offer this hypothesis as a likely, but not proven, explanation.This analysis suggests that the animals in the denser stocked, large ponds (105000 animals/ha) were experiencing periodic and chronic oxygen stress. The nature of this
278
A. Garcia 111,D. E. Brune
stress was such that it never contributed to a catastrophic die-off of animals. However, this caused the biomass to be regulated by the advective oxygen transfer, slowing the growth rate. These factors apparently led to a situation in which the animals were experiencing chronic stress and likely reduced the animal's resistant to opportunistic infection. SUMMARY Large 8 ha (20 acres) shrimp grow-out ponds stocked at rates in excess of 50 000 animals/ha are likely to be advection limited with respect to oxygen flux. This conclusion is supported by an array of circumstantial evidence consisting of observations by other researchers indicating a similar limit to stocking rates, the observation that large ponds are more prone to this limit than small ponds, calculations showing that diffusive mass transport is insufficient to meet oxygen demands at the level required in these ponds, and the observation of a 'biomass plateau' in heavily stocked ponds. Furthermore, calculated oxygen flux values suggest that large shrimp ponds will require 2 to 4 times the wind oxygenation capacity (to provide adequate mixing), while small ponds (0.1 ha) can operate successfully at a factor of 1-4 to 2.4 times excess oxygenation capacity. Paddlewheel power requirements (data taken from 0.25 ha ponds) are likely to range from 1.2 to 1.8 times oxygenation requirement to supply adequate pond mixing.
REFERENCES Banks, R. B. & Herrera, F. F. (1977). Effect of wind and rain on surface reaeration. J. Environ. Engng Div., 489-504. Broom, J. G. (1970). Shrimp culture. Proc. World Maricult. Soc., 1, 63-8. Chieng, C., Garcia, A. III, & Brune, D. E. (1988). Oxidation requirements of formulated micropulverized feed. J. WorldAquacult. Soc., 20, 24-9. Egusa, S. ( 1961). Studies on the Respiration of the 'Kuruma' prawn, Pennaeus japonecus. Bull. Japan. Soc. Sci. Fish., 27,650-9. Fast, A. W., Kent, E. L., Victor, J. E. & Gonzales, H. J. (1988). Effects of water depth and artificial mixing on dynamics of Phillipines brackish water shrimp ponds. Aquacultural Engineering, 7 (5), 349-61. Garcia, A. (1987). Unpublished data, Department of Agricultural Engineering, Texas A&M University, College Station, TX. Hanson, J. A. & Goodwin, H. L. (1977). Grow-out systems for peraeid shrimp. In Shrimp and Prawn Farming in the Western Hemisphere, ed., J. A. Hanson. Van Nostrand Co.
Oxygen transport in shrimp culture ponds
279
Hopkins, J. S., Baird, M. L. Grados, O. G. Maier, E E Sandifer, E A. & Stokes, A. D. (1989). Impacts of intensive shrimp culture practices on the culture pond ecology, unpublished manuscript obtained from the Waddell Mariculture Center, Blufton, South Carolina. Lewis, W. K. & Whitman, W. C. (1924). Principles of Gas Absorption. J. Ind. Eng., 16, 1215-20. Loehr, R. C. (1977). Pollution Control in Agriculture, Academic Press Inc., New York. Losordo, T. M. (1989). Department of Zoology, University of North Carolina, personal communication. Madenjian, C. E, Roger, G. L. & Fast, A. W. (1987). Predicting night time dissolved oxygen loss in prawn pond of Hawaii, Part I. A New Method, Aquacultural Engineering, 6, (3). 191-208. Parker, J. C., Conte, E S., Macgrath, W. S. & Miller, B. W. (1974). An intensive culture system for penaeid shrimp. Proc. WorldMaricult. Soc., 5, 65-79. Roger, G. L. & Fast, A. W. (1988). Potential effects of low energy water circulation in Hawaiian prawn ponds. Aquacultural Engineering, 7, (3), 155-66. Seidman, E. R. & Lawrence, A. L. (1985). Growth, feed digestibility and proximate body composition of juvenille Pennaeus vannamei and Pennaeus monodon grown at diffferent dissolved oxygen levels. J. World Aquacult. Soc., 16,333-346. Wong, B. R. (1983). Optimizacion de las Condiciones de Cultivo para Aumentar la Produccion di Camaron. Tesis de Centro de Investigacion Y de Estudios Avanzados Instituto Politecnico Nacional. Wyban, J. A., Lee, C. S., Sati, V. T., Sweeney, J. N. & Richards, W. K. (1987). Effect of stocking density on shrimp growth rates in manure fertilized ponds. Aquaculture, 61, 23-32.
Transport Limitation of Oxygen in Shrimp Culture Ponds A l b e r t G a r c i a III Agricultural Engineering Department, Texas A&M University, College Station. Texas 77843, USA
D a v i d E. B r u n e Agricultural & Biological Engineering Department, Clemson University,Clemson, South Carolina 29634, USA (Received 9 January 1989; accepted 11 July 1991 ) ABSTRACT A 120-acre shrimp farm located in southeast Texas has failed for three years to produce expected yields of shrimp despite intensive management. Data obtained from these ponds are combined with observations from other culture facilities to support the hypothesis that the degree of mixing of pond water is the primary limitation to oxygenation of the benthic biomass in a shrimp culture pond. The implication of this conclusion is that shrimp ponds are advection limited," therefore, rational design of such facilities will require detailed data concerning the impact of interactions of pond size, depth and geometry with wind speed, and aerator placement, upon pond mixing. INTRODUCTION A set of six, 8 ha (20 acre), ponds approximately 1 m (3-28 ft) deep located on the Texas coast south of Corpus Christi, Texas, were stocked with Penaeus vannamei at different times and stocking densities. Ponds 1 through 4 were stocked on 26 May, 1987 with 0-62 g animals at approximately 105 000/ha ( - 10/mZ). Ponds 5 and 6 were stocked on 30 April, 1987 with 0"8g animals at 4 5 0 0 0 / h a ( - 5 / m 2 ) . Growth proceeded normally in ponds 5 and 6, but was retarded in ponds 1 to 4 (see Table 1 ). 269 Aquacultural Engineering 0144-8609/92/S05.00 - © 1992 Elsevier Science Publishers Ltd, England. Printed in Great Britain
270
A. Garcia Ill, D. E. B r u n e
After 5 weeks, ponds 1 through 4 experienced a significant amount of bird predation which is indicative of excessive mortality. Mortality in these ponds was estimated to be one percent per day. At this time the average shrimp weight in ponds 1-4 was 3"5 g. Ponds 5 and 6 were harvested after 11 weeks. The average weight in these ponds was 13"5 g and the total yield was 540 kg/ha. Oxygen concentration and temperature, which were monitored twice daily, did not indicate periods of extreme stress. The pathology of the animals in ponds 1 through 4 during the growout period did not indicate a singular cause of death. The data showed that the animals contained an increasing amount of opportunistic microflora as time progressed. While it is impossible to ascertain with certainty the cause of mortality in ponds 1 through 4, it is possible to collect facts and hypothesize a likely explanation. The following analysis is presented as an attempt to describe the mechanism for mortality in this particular case, and to propose general research objectives to address the suggested problem. BACKGROUND
Pond oxygen dynamics Oxygen management in aquaculture ponds is made difficult not only by the complex web of ecological interactions but also by diurnal fluctuations. During periods of light, the photosynthetic biomass both produces and consumes oxygen. Usually the production exceeds the light period respiration thus elevating the pond oxygen levels. On calm days, intensive algal bloom often elevates the water oxygen concentration to superTABLE ! Growth (g) Data for Ponds Stocked at 45 000" and 105 000 t' Animals Elapsed time
P o n d 5"
l ' o n d 6"
Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9 Week 10
3.17 4.50 5.23 5-70 6.50 5"90 8.40 9.20 10'9
3.70 5.00 6.23 5.80 6-50 7.80 9-20 9.40 10'0
l ' o n d I t,
t ' o n d 2 I'
Pond 3 h
l'ond 4 h
4-06 5-20 3.50
4-30 4-80 3.40
4"25 4.96 3"40
4'10 4"40 3"80
Oxygen transport in shrimp culture ponds
271
saturated concentrations. During the night, metabolism shifts completely to respiration. The oxygen may be rapidly consumed, with the lowest oxygen levels normally occurring just before dawn. Whenever pond 02 concentrations are below saturation, wind will serve to reaerate the pond with the rate of mass transfer of oxygen a function of wind speed. However, when the pond is at supersaturated conditions the wind will act to de-oxygenate the water. In addition to the diurnal variation in oxygen levels, there is an ever present biochemical oxygen demand (BOD) from added shrimp feed. BOD is exerted in the pond in two zones, one part associated with the water column and a second part associated with the sediment (SOD). The SOD is usually the more significant component affecting the pond water quality in ponds rearing benthic organisms at productivities in excess of 1000 lb/acre (Fast et al, 1988). The sources of the organics driving the SOD are the addition of artificial feeds and fertilizers, influent water and decay of sinking autotrophic biomass. For simplicity of analysis one may consider the sum of animal respiration and the respiration of the fecal matter to be equal to the ultimate oxygen demand of the feed added to the pond. From the perspective of the shrimp, the water near the sediment is the environment in which they must feed and grow, and the water column represents the reservoir of oxygen. In a well mixed and adequately aerated pond, the two are the same and no distinction is necessary. In such cases, advective oxygen mass transfer is sufficient for benthic demand. A typical management practice is to assume the pond is well mixed and to measure the oxygen level at the mid depth in the late afternoon and early morning and assume that such a measurement represents the environment to which the shrimp are exposed.
Oxygen transport limitations A problem may arise when the assumption that the pond is well mixed is no longer valid. During a calm period, advective mass transport may be virtually eliminated or, depending on depth and pond size, advective transfer may be limited to a shallow zone, even during windy days, and may not provide sufficient oxygen transfer, particularly at high densities. At such time, the only mechanism for mass transfer will be diffusion. Under such conditions, the diffusivity of oxygen in water becomes the limiting factor to oxygen transport to the shrimp biomass. A simple way of looking at this is to focus on oxygen flux. Oxygen enters the water at the air-water interface at a rate described by the two-film gas transfer theory (Lewis & Whitman, 1924). It is partially consumed/produced in
272
A. Garcia 11I, D. E. Brune
the water column and diffuses to the sediment. If the SOD, oxygen differential across the diffusive zone, the diffusivity of oxygen in salt water, and the depth of the diffusive zone are known, then the rate of oxygen supply to the sediments can be calculated. Normally it is the reverse calculation that is more revealing, which can be stated as; what is the maximum thickness of an unmixed layer of water that can deliver enough oxygen to the water-sediment interface to meet the shrimp and SOD requirements while maintaining a minimum oxygen level of 0-5 ppm (lethal limit)? To answer this, the SOD must be estimated and a best case may be constructed for the oxygen levels present in the pond. It is known from previous work (Chieng et al., 1988) that the ultimate BOD of the shrimp feed is 650 mg 02/g. At a stocking rate of 45 000 and 105 000 animals/ha, and an overall feeding rate ranging from 1% to 7% of body weight per day, the SOD may be calculated (Table 2). Measured values of SOD, in these ponds, were seen to be in the range of these calculated values (Garcia, 1987). In calculating the 'best case' for diffusive transport of oxygen at the pond sediment interface, the oxygen in the bulk pond water was assumed to be near saturation (8.0 ppm). This would likely represent the oxygen condition on a bright calm day. A level of 0.5 ppm at the sediment-water interface was chosen as the lethal limit to sustain P. vannamei. The thickness of a hypothetical diffusive layer at the sediment water interface may be calculated as follows (see Fig. 1 ):
TABLE 2 Calculated Benthic Oxygen Uptake Rate of Shrimp Ponds (g O2/m -~ day) at Feeding Rates Varying from 7% to 1% of Body Weight per Day Animal weight (g)
1 5 10 15 20 25
Number of animals/m:
Feed rate (%)
5
lO
20
40
lO0
200
0"23 0"82 0"82 0-74 0-98 0-82
0"46 1"64 1'64 1"48 1"96 1'64
0-92 3-28 3'28 2-96 3"92 3"28
1"84 6"56 6'56 5"92 7"84 6"56
4-60 16"40 16"40 14-80 19"60 16'40
9"20 32"80 32"80 29"60 39"20 32"80
7 5 2'5 1"5 1"5 1'0
Oxygen transport in shrimp cultureponds
273
DO m 8.0 m9]l (Saturation)
ADVECTIVE ZONE
~ Z== thloknou
i
/)
.T.
oo = o.s mo/I
DIFFUSIVE
POND S E D I M E N T -
Fig. 1.
ZONE.
,
SOD :
19 m 9 0 l / m 2 / h
C~,,h,Um,O
Representation of diffussive oxygen transport in a shrimp pond.
where F--oxygen flux, mg OE/m 2 day; D=oxygen diffusivity, 2.07 × 10 -4 mE/day; Z = thickness of diffusive zone, m; ( d c / d z ) = c o n centration gradient, mg/m 3. In the case of a pond stocked with 1 g animals at 10 animals/m 2 (an estimated SOD of 460 mg OE/m 2 day) this calculation suggests a diffusive layer limit of 0-34 cm. This is a highly significant statement, indicating the necessity of maintaining mixed conditions in a pond. Even under the most favorable conditions at stocking densities of 105 000 animals/ha (10/mE), a stagnant layer of 0-34 cm would be the limit at which oxygen could diffuse to the sediment rapidly enough to meet the lowest oxygen demand experienced during the 190 day growout period. Therefore, under most conditions, the pond bottom and hence the shrimp are totally dependant upon advective transport for mixing (and 02 transport) to within less than 0.5 cm of the bottom. Response of shrimp to oxygen concentration Previous investigations of penaeid shrimp oxygen response have suggested that the respiratory response of these animals is dependant on a number of factors including oxygen level, animal size, activity and species. For smaller animals (less than 3-0 g) respiration per unit biomass can be more than twice as high than that of larger animals (Egusa, 1961 ). Egusa's data show that large shrimp (greater than 5.0 g) are probably respiratory regulated at oxygen levels of 2.0 rag/liter. Data from Seidman
274
A. Garcia 111, D. E. Brune
and Lawrence (1985) confirm the observation that 2.0 mg/liter is a concentration that will limit growth. It is not unreasonable to assume that concentrations of 2.0 mg/liter may lead to animals stress; this assumption is supported by the observation that raceway culturists prefer to maintain outlet dissolved oxygen levels in excess of 3.5 mg/hter (Wong, 1983). This information suggests that the shrimp respiration in the observed ponds was under constant regulation by the pond oxygen level. Further, if the oxygen flux to the sediment is subject to the degree of mixing of the water column (as suggested by the previous calculation), then it is reasonable to hypothesize that the degree of animal stress (and consequently animal mortality) is also linked to the intensity of pond mixing at increasing carrying capacity. The actual pond culture involves larger shrimp, more sediment oxygen demand, nighttime respiration and lower oxygen levels in the water column. These conditions will likely lead to the imposition of chronic stress, if not, outright catastrophic failure.
DISCUSSION If the degree of pond mixing is limiting shrimp biomass yields from culture ponds, then one would expect that the required horsepower to sustain high density shrimp cultures would be greater than that needed to simply oxygenate the pond. For example, a typical waste treatment lagoon being loaded at a rate of 50 lb (22.7 kg) of BOD/acre day would require approximately 2 hp/acre (0.6 kw/ha) to adequately aerate the lagoon (Loehr, 1977). However, as Loehr observes, such a pond would require nearly 29 hp/acre (8.8 kw/ha) to keep the bacterial and waste solids in suspension. Therefore, there exists an approximately 10-fold difference in power requirements to mix rapidly settling particles as opposed to air-water interfacial oxygen transfer. The logical question to ask is : Does a similar situation exist for shrimp culture ponds; in other words, what is the increased horsepower requirement to maintain a mixed zone at the pond bottom? Table 3 gives the oxygenation rates expected from paddlewheel aerators operating at power levels observed by Hopkins et al. (1989 unpublished) to provide 'adequate' pond oxygen conditions (0.1 hp/animal m2). Such aerators will typically provide between 1.8 and 2.0 lb of O2/hp-h hr (1-2 kg/kw-h) at an oxygen concentration of 0 mg/liter. In Table 3 the rates are adjusted for 3 mg/liter and 5 mg/liter pond 02 levels and 12 and 24 hr aerator operation cycles.
Oxygen transport in shrimp culture ponds
275
TABLE 3 Aerator oxygenation (g O2/m ~ day) at Aerator Power Levels Suggested by Hopkins et al. (1989)
Aerator cycle (h)
Oxygen level (mg/liter)
5
10
20
40
100
2(X)
12
5 3 5 3
0'43 0-68 0"86 1-36
0"86 1-36 1"72 2"72
1"72 2'72 3.44 5'44
3"44 5'44 6"88 10-88
8"60 13"60 17"20 27'20
17'20 27'20 34.40 54.40
0.5
1.0
2.0
4"0
10.0
20.0
24
Recommended aerator power (hp/acre)
Number of animals/m-'
In contrast, wind driven surface re-aeration at a wind speed of 5.4 m/s (average annual wind for the southeast Texas coast) are predicted by equations from Banks and Henera (1977) to be 3.7 gm/m 2 day at 3.0 rag/ liter of pond dissolved oxygen or 2.3 g / m 2 day at 5.0 mg/liter of pond dissolved oxygen levels. Ponds 5 and 6 (see Table 1 ) were unaerated and successfully produced 10 g animals stocked at 5 animals/m 2. At this size and density, Table 2 predicts a maximum SOD of 0.98 g/m 2 day with a wind reaeration potential of 2"3 to 3.7 g / m 2 day or a ratio of (potential/required) aeration of 2"3 to 3.7. In the unsuccessful ponds (1-4) this ratio drops to 1-2-1.8. This suggests that approximately two to four times as much power (supplied by wind) is needed to maintain a 'healthy' pond as would be expected based on aeration requirements alone. In the higher density ponds (20-200 animals/m 2) which are always aerated, the ratio averages 1.4. Again, more power is needed to maintain the pond than is expected (based on aeration requirements) however, the ratio is less, probably because paddlewheel aerators are more effective at mixing the diffusive layer than surface wind action. Figure 2 presents the calculated average shrimp respiration rates in ponds 1-4 and ponds 5 and 6. This figure further supports the oxygen limitation concept. Shrimp respiration represents approximately 10-20% of total feed respiration (calculated from data presented by Chieng et al., (1988) Seidman and Lawrence (1985) and Egusa (1961). Figure 2 graphically demonstrates how the shrimp biomass can continue to expand as long as total pond respiration is at least a factor of 2.0 below pond reaeration (supplied by wind). However, pond 5 and 6 peaked early above this value; the end result being a slow shrimp mortality as the
276
A. Garcia Ill, D. E. Brune
500-
~
400-
PONDS !, 2, 3, 4
>-
300. I
.......
E c~ 200-
E
100-
0 0
Fig. 2.
///
PONDS 5, 6
......... I ......... I ......... t ......... [ 5 10 15 20
TIMES (Weeks) Shrimp oxygen uptake rates in successful (5 and 6) and unsuccessful ( 1,2, 3 and 4) grow-out ponds.
biomass 'self corrects'. If aquatic biomass accumulation is limited by oxygen then an expected response would be for the total mass of shrimp to achieve a maximum limit regardless of stocking density. This phenomena has been demonstrated previously with P. vannamei (Wyban et aL, 1987). This effect can manifest itself in two ways. The first is for growth to continue as normal for part of the population. As the biomass attempts to expand, a certain portion must die in order to prevent the total 02 demand from surpassing the biomass limit. Another manifestation of the same problem is for growth to cease altogether; with the total biomass remaining at maintenance levels. The actual response is usually some combination of the two possibilities. As early as 1970, Broom et al. speculated that the shrimp stocked in large unaerated ponds in excess of 50 000 animals/ha would be limited by the oxygenation capacity of the pond. He also concluded that higher stocking densities and yields would be possible when economical aeration systems were developed. Tables 2 and 3 suggest that this limit is, in fact, 50 to 25% of the true oxygenation capacity, and more closely represents the mixing limit of the pond as a result of wind action. This conclusion is supported by a study conducted by Rogers and Fast (1988) in which they found that freshwater prawn production could be
Oxygen transport in shrimp culture ponds
277
increased 8-11%, at 1600 lb/acre (1800 kg/ha) with induced mixing of less than 0"25 hp acre. Past researchers (Parker et al., 1974) have observed that stocking rates up to 80000 animals/ha (8-0 animal/m 2) in small unaerated, or minimally aerated, 0"1 ha ponds does not produce significant mortality. This suggests that small ponds are better mixed than large ponds. Wind action on small ponds may result in increased mixing as water is pushed against the bank and circulated to deeper layers. In spite of the fact that larger ponds (with a larger fetch) produce greater wind-induced wave height, this apparently does not produce greater bottom water mixing in the pond. It is common practice to position shrimp ponds with short axis into the prevailing wind direction (Losordo, 1989) or at least, to turn the long axis 30 degrees away from the wind (Hanson & Goodwin, 1977). This implies that shore action is necessary for proper mixing; shore length per unit of water volume may be a better indication of pond mixing than pond fetch. The case of 8.0 animal/m 2 represents a maximum SOD of approximately 1.57 g/m 2 day. This suggests a potential/required aeration ratio ranging from 1.46 to 2"36 which suggests that the efficiency of wind mixing in small (0-1 ha) ponds approaches the level of paddlewheel mixing in larger ponds. The efficiency of wind induced mixing in large 8 ha will likely be significantly reduced. In a recent related study, Wyban et al. (1989) demonstrated successful culture of Penaeus vannamei at 25 animal/m 2 in unaerated ponds located at Makapuu Point, Hawaii. The average wind speed at this location (during the growing season) is approximately 16% greater than Corpus Christi, Texas giving an overall wind aeration rate (at 5.0 rag/liter) of 4.48 g / m 2 day as opposed to 3"68 at Texas. At 25 animal/m 2 this gives aeration potential/required ratio of 1"2 which is at the same level which produced failure in the Texas ponds. However, the Texas ponds were 8 ha (20 acre) while the Hawaiian ponds were only 0.4 ha (1 acre). The observation of increased wind-induced oxygenation of Hawaiian ponds is supported by Madenjian et al., (1987). Interestingly, Wyban et al. (1987) found that in the unaerated ponds, individual animal weight had only reached 15 g/animal at a ratio of 1.2 while similar, aerated ponds expanded to 21 g/animal, at which time the ratio again approached 1.25. Due to the very limited data base (as usually is the case) any reasons proposed for the observed Texas shrimp mortality are speculative at best. The authors are well aware of this and offer this hypothesis as a likely, but not proven, explanation.This analysis suggests that the animals in the denser stocked, large ponds (105000 animals/ha) were experiencing periodic and chronic oxygen stress. The nature of this
278
A. Garcia 111,D. E. Brune
stress was such that it never contributed to a catastrophic die-off of animals. However, this caused the biomass to be regulated by the advective oxygen transfer, slowing the growth rate. These factors apparently led to a situation in which the animals were experiencing chronic stress and likely reduced the animal's resistant to opportunistic infection. SUMMARY Large 8 ha (20 acres) shrimp grow-out ponds stocked at rates in excess of 50 000 animals/ha are likely to be advection limited with respect to oxygen flux. This conclusion is supported by an array of circumstantial evidence consisting of observations by other researchers indicating a similar limit to stocking rates, the observation that large ponds are more prone to this limit than small ponds, calculations showing that diffusive mass transport is insufficient to meet oxygen demands at the level required in these ponds, and the observation of a 'biomass plateau' in heavily stocked ponds. Furthermore, calculated oxygen flux values suggest that large shrimp ponds will require 2 to 4 times the wind oxygenation capacity (to provide adequate mixing), while small ponds (0.1 ha) can operate successfully at a factor of 1-4 to 2.4 times excess oxygenation capacity. Paddlewheel power requirements (data taken from 0.25 ha ponds) are likely to range from 1.2 to 1.8 times oxygenation requirement to supply adequate pond mixing.
REFERENCES Banks, R. B. & Herrera, F. F. (1977). Effect of wind and rain on surface reaeration. J. Environ. Engng Div., 489-504. Broom, J. G. (1970). Shrimp culture. Proc. World Maricult. Soc., 1, 63-8. Chieng, C., Garcia, A. III, & Brune, D. E. (1988). Oxidation requirements of formulated micropulverized feed. J. WorldAquacult. Soc., 20, 24-9. Egusa, S. ( 1961). Studies on the Respiration of the 'Kuruma' prawn, Pennaeus japonecus. Bull. Japan. Soc. Sci. Fish., 27,650-9. Fast, A. W., Kent, E. L., Victor, J. E. & Gonzales, H. J. (1988). Effects of water depth and artificial mixing on dynamics of Phillipines brackish water shrimp ponds. Aquacultural Engineering, 7 (5), 349-61. Garcia, A. (1987). Unpublished data, Department of Agricultural Engineering, Texas A&M University, College Station, TX. Hanson, J. A. & Goodwin, H. L. (1977). Grow-out systems for peraeid shrimp. In Shrimp and Prawn Farming in the Western Hemisphere, ed., J. A. Hanson. Van Nostrand Co.
Oxygen transport in shrimp culture ponds
279
Hopkins, J. S., Baird, M. L. Grados, O. G. Maier, E E Sandifer, E A. & Stokes, A. D. (1989). Impacts of intensive shrimp culture practices on the culture pond ecology, unpublished manuscript obtained from the Waddell Mariculture Center, Blufton, South Carolina. Lewis, W. K. & Whitman, W. C. (1924). Principles of Gas Absorption. J. Ind. Eng., 16, 1215-20. Loehr, R. C. (1977). Pollution Control in Agriculture, Academic Press Inc., New York. Losordo, T. M. (1989). Department of Zoology, University of North Carolina, personal communication. Madenjian, C. E, Roger, G. L. & Fast, A. W. (1987). Predicting night time dissolved oxygen loss in prawn pond of Hawaii, Part I. A New Method, Aquacultural Engineering, 6, (3). 191-208. Parker, J. C., Conte, E S., Macgrath, W. S. & Miller, B. W. (1974). An intensive culture system for penaeid shrimp. Proc. WorldMaricult. Soc., 5, 65-79. Roger, G. L. & Fast, A. W. (1988). Potential effects of low energy water circulation in Hawaiian prawn ponds. Aquacultural Engineering, 7, (3), 155-66. Seidman, E. R. & Lawrence, A. L. (1985). Growth, feed digestibility and proximate body composition of juvenille Pennaeus vannamei and Pennaeus monodon grown at diffferent dissolved oxygen levels. J. World Aquacult. Soc., 16,333-346. Wong, B. R. (1983). Optimizacion de las Condiciones de Cultivo para Aumentar la Produccion di Camaron. Tesis de Centro de Investigacion Y de Estudios Avanzados Instituto Politecnico Nacional. Wyban, J. A., Lee, C. S., Sati, V. T., Sweeney, J. N. & Richards, W. K. (1987). Effect of stocking density on shrimp growth rates in manure fertilized ponds. Aquaculture, 61, 23-32.