Chiller water treatment from channel catfish (Ictalurus punctatus) processing plants with ultrasound, ozone, and pulsed-light.

High-frequency ultrasound, ozone, and pulsed-light were studied as alternative disinfecting treatments for catfish chiller water. Chlorination of process water is done on site, by processors, or by municipalities. However, contaminated water may reach the product/process if water is not adequately treated. Incoming quality (both fresh and recycled), and discharge water streams of two channel catfish processing plants was determined. Quality of incoming fresh water was better than recycled chiller water, which was similar to discharge water. High-frequency ultrasound (850 MHz) had no effect on microbial reduction of chiller water. Ozone at 7.68 ppm and 5 min contact time killed Listeria spp. and reduced heterotrophs and coliforms by 1 log CFU/mL. Of the Gram-negative microorganisms, Acinetobacter baumannii was found to be one of the most resistant to ozone at low concentrations (3 ppm). Pulsed white light reduced heterotrophs and coliform counts by more than 4 log CFU/mL in recycled chiller water. Thus, pulsed -light showed the best results for treatment of catfish processing water, followed by ozone (> 5 ppm) while high-frequency ultrasound had no effect.

KEY WORDS: Water, quality, chiller, catfish, disinfection 

**********

Channel catfish (Ictalurus punctatus) is one of the most popular/eaten finfish in the United States. Most channel catfish production takes place in the southeastern part (Mississippi and adjoining states) of the United States (TCI, 2001). The water used in the chiller, to lower the temperature of the fish, must "be safe and of adequate sanitary quality" (21 CFR 110.37, FR 1995). The quality of the water determines the overall quality of the product itself during processing. The chiller is used to cool the end product and may aid in surface cleaning (dilution) before product is packaged. In the southeastern part of the United States (SE U.S.), rain recharges the aquifers. Also, the SE U.S. has an abundance of poultry, cattle, hogs, sheep, catfish, and horse farms. These animals may carry many microorganisms, which are discharged in their feces and become part of the agricultural run-offs that pollute the water (Gray, 1994). The temperature of the water in the chiller (< 5[degrees]C) must be monitored and contr olled at all times. The chiller is usually the final wash that the catfish goes through before packaging, so the water in the chiller should be as sanitary as possible.

There are many new uses of ultrasound and more to come in the near future. Some of the uses of ultrasound now include acoustical agglomeration and treatment of water and wastewater (Crum, 1992; Mason, 1993). Phull et al. (1997) conducted an experiment on the effects of ultrasound on the biocidal treatment of water and its effect on microorganisms such as Eschericia coli. The results showed that ultrasound could be used to kill microorganims. Furthermore, high frequency ultrasound was suggested to be more beneficial than low frequency ultrasound.

Ozone has been used in potable water treatment since 1856 (Langlais, 1990). The United States Food and Drug Administration (USFDA) has authorized the use of ozone as Generally Recognized as Safe (GRAS) (66 FR 33829, 2001). Ozone is now used in the bottled water industry, soft drink industry, and beer and wine industry. In the bottled water industry, ozone is applied to water just before bottling.

The FDA has approved pulsed light (PureBright[R] Technologies, San Diego, CA) to reduce microbial contamination in water. This technology works by storing electrical energy in a capacitor and releasing it in short, high intensity light pulses (Barbosa-Canovas et al., 1998). These pulses cause a lamp to produce short light flashes (< 1 [micro]s) and these short-duration flashes inactivate microorganisms. The PureBright[R] process uses a pulsed power system to drive an inert gas lamp that emits broadband light pulses approximately 20,000 times more intense than sunlight. PureBright[R] is comprised of light wavelengths between 200 and 300 nm in the ultraviolet (UV), visible, and near infrared (IR) parts of the spectrum. The process produces no significant heat rise in the product (PurePulse Technologies, 2002; Figueroa et al., 2002).

The objectives of this research were to explore disinfection of chiller water by alternate technologies, in order to recycle it and decrease the likelihood of microbial cross-contamination. The specific objectives included: (1) evaluating water and recycled chiller water quality of two catfish processing plants, (2) studying the possible effect of ultrasound and ozone alone and in combination in disinfecting chiller water, and (3) studying the effect of pulsed-light technology on microbial load and waterborne pathogens of chiller water.

MATERIALS AND METHODS

Quality of Chiller Water--Catfish chiller water was collected in the summer and fall 2000 from two different catfish processing plants in Mississippi. One plant was located in northwest Mississippi (NW) and the other one was located in northeast Mississippi (NE). Water from the chiller was sampled. For the NW plant, the incoming chilled water was "fresh" or "clean" water, whereas for the NE plant, the incoming water was recycled water. Chiller discharge water was also sampled at both plants. For both, incoming and discharge water, samples were taken every 15 min for a period of 45 min.

Samples were placed in sterile bottles and packed with ice in an ice chest for transportation. Upon arrival at the laboratory, a composite sample was prepared by mixing samples taken at different time intervals. The physical and chemical characteristics of the chiller water were analyzed: pH, turbidity, hardness, total solids, and settleable solids by the method of APHA (1998). Total coliform counts were also performed as suggested by APHA (1992).

Disinfection Trials--An ultrasonic K80 generator (Meinhardt, Germany), with 100 W maximum power output was used in combination with a static ultrasonic chamber. The reaction chamber has 346 mL of useful volume. A water column of 2.0 cm corresponding to 77 mL was used in this section (Chamul, 2000). Constant for each treatment was a frequency of 850 MHz and amplitude of 580 mV. Contact times were 0, 10, and 15 min. Discharged recycled chiller water from the northeast Mississippi plant was used in this part of the study.

Ozone was produced from purified, extra dry oxygen by using a Welsbach Model T-816 Ozonator (Welsbach Ozone Systems Corporation, Philadelphia, PA) set at 60 Hz and 29.58 kPa (8 psig) dry oxygen pressure. The flow of ozone was set at 2 L/min (S.L.P.M.). The produced ozone gas was dispersed through a diffuser (average pore size 60 [micro]m, Fisher Scientific, Houston, TX) into the chiller water for 30 sec to identify possible resistant bacteria to ozone, and for 0, 5, or 10 min for the disinfection trial. Ozone concentration was measured using an iodometric method (Stephens, 1984) and results were expressed as mg/L or ppm [O.sub.3]. The surviving Gram-negative microorganisms were isolated on Eosin Methylene Blue Agar. After isolation, API 20E biochemical tests, from bioMerieux (Hazelwood, MO) were performed in order to identify the surviving microorganisms.

Chiller water from the catfish processing plant in northeast Mississippi was collected in 18 L (5 gallon) buckets, transported to the laboratory at room temperature and treated within 2 h. Water was passed through a PureBright[C] Water Purification System Model PBW-4 (Maxwell Technologies, San Diego, CA) and samples collected once steady flow was achieved. The flow rate was 2.21 L/min (0.584 gal/min). For the detection of Listeria spp. (tested on disinfection trials with ozone and ultrasound), the TECRA Listeria Visual Immunoassay Kit (TECRA DIAGNOSTICS, Australia) was used. The water sample was added to 225 mL of Modified Listeria Enrichment Broth (LEB) (Difco Laboratories, Detroit, MI) and incubated for 24 h at 30[degrees]C. After 24 h, 0.1 mL of the LEB culture was transferred to a tube of Fraser Broth (Difco Laboratories, Detroit, MI). Before transferring, 50 [micro]L/10 mL 0.5% acriflavin, 40 [micro]L/10 mL 0.5% nalidixic acid in 0.1 M NaOH, and 0.1 mL 5% ferric ammonium citrate were added to the Fraser broth. The tubes were then incubated for 22-24 h at 30[degrees]C. After incubation, 1 mL was transferred from the enrichment broth to a labeled tube and then the TECRA kit was used. Results were read and expressed as positive or negative.

Statistical Design--The survey of the chiller water was set as a randomized complete block (RCB) design with two factors: plant (2) and sampling month (4), with 2 replications (blocks). The replicates were composite from three samples taken every 15 minutes for 45 minutes. Analysis of data was performed using PROC ANOVA (SAS, 1997). When there was a significant difference (P [less than or equal to] 0.05), means were separated by Fisher's protected LSD (Dowdy and Wearden, 1985).

RESULTS AND DISCUSSION

Quality of Chiller Water--For the incoming chiller water, there was a difference (P [less than or equal to] 0.05) in temperature by sampling month (data not shown). For the plant in NW Mississippi, the temperature of water in the chiller ranged from 8[degrees]C to 12[degrees]C during the sampling period. The temperatures for the water in the chiller in the plant in NE Mississippi ranged form 6[degrees]C to 9[degrees]C. Ambient temperatures ranged from 31.1 to 31.4[degrees]C in west Mississippi and from 31.4 to 30.3[degrees]C in east Mississippi. The ambient temperature affects the temperature of the ground water. Although the water is being chilled before it goes into the chiller, the outside temperature could still affect the temperature of the water going into the chiller. If warmer water is coming into the plant, it will take more energy to chill that water to the desired temperature.

Means for the physical and chemical parameters of the incoming water, by plant location, are summarized in Table 1. The mean pH values were 7.3 and 6.9 for fresh and recycled water, respectively. These pH values fall into the range (6.5-8.5) of the primary standards set by the EPA (2001). Turbidity for fresh water ranged from 0.066-0.300 NTU, with a mean value of 0.14. For recycled water the turbidity values were higher (P [less than or equal to] 0.05) ranging from 14.93-51.30 NTU, with a mean value of 28.71, these values are higher than the maximum level (5 NTU) stated by the World Health Organization (WHO, 1985) for processing water. Because turbidity is a measure of the presence of suspended solids, it may be an indication of concentration of microorganisms present (McGhee, 1991). Therefore, the higher the turbidity, the higher the microbial load. The recycled water had higher turbidity (P 0.05) and a higher (P [less than or equal to] 0.05) and a higher (P [less than or equal to] 0.05) bacterial count. Th is indicates that the water was not being properly treated (Gray, 1994) prior to recycling. Water hardness was lower (P [less than or equal to] 0.05) for fresh water (0.18 mg [CaCO3.sub.3]/L) than for recycled water (0.35 mg [CaCO.sub.3]/L). Total solids were higher (P [less than or equal to] 0.05) for recycled water (265-635 mg/L) than fresh water (160-350 mg/L).

There was a significant difference in heterotrophic plate counts (HPC) and total coliform counts (TCC) by plant (Table 1). Fresh water had an HPC density of 0.2 log CFU/mL while the recycled water had 4.4 log CFU/mL. This difference in microbial load could be attributed to the fact that the plant in east Mississippi recycles the water used in the chiller. Chamul (2000) found that the microbial load of the ground water in east Mississippi had values that ranged from 0.7 to 3.2 log CFU/mL for HPC and 0.0 to 3.0 log CFU/mL for TCC. The quality of the ground water for the plant in east Mississippi fell into the normal range of HPC, which may vary from 1 to 5 log CFU/mL. Once more, the recycled water had a higher microbial count than the fresh water. Total Coliform counts were 0.1 and 3.4 log CFU/mL for fresh and recycled water, respectively. These levels exceeded those of the EPA (2001), thus calling for better monitoring and treatment of raw (incoming) water.

Temperature of discharge water in the chiller (Table 2) ranged from 8[degrees]C for fresh water to 10[degrees]C for the recycled discharge water. The tendencies in temperature were about the same for both plants, but the temperature was always higher for the plant in NE Mississippi. Means for the physical and chemical parameters of the outgoing chiller water, by plant location, are summarized in Table 2. There was a significant difference in turbidity between plants, with mean values of 16.6 NTU for fresh water and 33.7 NTU for recycled water. For both plants, the water leaving the chiller had higher turbidity values than the incoming or recycled water. This is expected because the water "picks-up" solids from the fish. This water is now considered to be wastewater. Water hardness was 0.23 mg [CaCo.sub.3]/L for fresh water and 0.44 mg [CaCo.sub.3]/L for recycled water. Settleable solids were 0 mg/L for recycled water and 0.63 mg/L for fresh water. There was also a significant difference in total solids, with values of 503 mg/L for recycled water and 990 mg/L for fresh water.

Both HPC and TCC were higher (P [less than or equal to] 0.05) for the recycled water (at discharge) by about 0.7 to 0.8 log CFU/mL compared to the fresh water at discharge (Table 2). Nunez (1995) and Fernandes et al. (1997) have shown that summer fish have higher TCC than in other seasons. Therefore temperatures higher than 4[degrees]C should be avoided during the chilling process.

Disinfection Trials--Recycled chiller water was exposed to ozone to discern its antimicrobial effects. Survival of Gram-negative microorganisms was also isolated and identified. The most common identified surviving microorganisms, after 30 sec of treatment (no residual ozone) of chiller water, was Acinetobacter baumannii (Table 3). Kim et al. (2000) reported that Acinetobacter spp. in channel caffish fillets was resistant to ozone and hydrogen peroxide treatments. Other resistant bacteria identified were Aeromonas hydrophilia and Providencia alcalifaciens (Table 3) which have been reported as normal flora in catfish processing plants (Wachalatone, 1996). Exposure of chiller water to over 60 sec (6.0 ppm [O.sub.3]) and 120 sec (7.0 ppm [O.sub.3]) resulted in no surviving microorganisms.

Samples of chiller water from a catfish processing plant were treated with ultrasound, ozone, and a combination of ultrasound and ozone. The results from this experiment are summarized in Table 4. All samples of both incoming and outgoing water were positive for Listeria spp. High frequency ultrasound (HFU) was not effective in reducing heterotrophic plate count (HPC), total coliform count (TCC), or Listeria spp. Ozone killed Listeria spp. and reduced HPC and TCC by about 1 log CFU/mL. There was no synergistic effect between HFU and ozone. The results of ozone treatment indicated that a longer contact time is needed to reduce the number of TCC, than the time needed to reduce the number of HPC. According to Langlais et al. (1990), Eschericha ccli is one of the most sensitive organisms to ozone treatment, while the Gram-positive cocci, Grampositive bacillae, and the mycobacteria are the most resistant.

Catfish chiller water samples were passed through the PureBright[R] Water Purification System. A flow rate of 2.21 L/min (0.58 gal/min) was used. At this flow rate the microbial load was reduced from 4 to 0 log CFU/mL (Table 5). Several authors have reported successful reduction microbial populations with light pulses (BarbosaCanovas et al., 1998; MacGregor et al., 1998; Rowan et al., 1999), reporting reductions of cell populations of 6 and 7 log orders in E. coli 0157:H7 and Listeria monocytogenes.

The pH of the water, after treatment, remained the same value as the pH of the water before treatment (Table 5). The pH values ranged from 6.8 before treatment to 7.4 after treatment. The turbidity of the chiller water decreased as it passed through the system, from 20.3 NTU to 10 NTU after treatment (Table 5). One of the specifications for the PureBright[R] Water Purification System is that the turbidity of the water leaving the system will be below 10 NTU. This is because high turbidity will impede penetration of the PureBright[R], a combination of UV, visible, and infrared spectrum.

CONCLUSIONS

The water used in catfish processing plants in Mississippi is of good quality, but if a processing plant intends on recycling the water in the chiller, a secondary treatment may be necessary. After evaluation of the chiller water samples, it was concluded that all samples were positive for Listeria spp. High frequency ultrasound was not effective in reducing the heterotrophic plate count (HPC), total coliform count (TCC), or Listeria spp.

Ozone reduced HPC and TCC by 1 log, but killed Listeria spp. Surviving microorganisms from chiller water samples exposed to ozone were isolated and identified. The predominant Gram-negative microorganism was Acinetobacter baumannii.

Using the PureBright[C] Water Purification System, with a flow rate of 2.21 L/min, the microbial load HPC and TCC was reduced from 4 log CFU/mL to 0. The turbidity of the chiller water was lower after passing through the PureBright[R] System. Ozone and PureBright[C], because of the antimicrobial properties they possess, could both be a secondary treatment in a catfish processing plant, if recycling is an option. These technologies may be accompanied by settling and filtration to reduce suspended solids, thus increasing treatment effectiveness.

Table 1

Mean of all parameters of the incoming freshwater and recycled chiller
water from two catfish processing plants.

Parameter                          Water Source        LSD (1)  C.V. (2)
                                 Fresh    Recycled

Temperature ([degrees]C)        9.5 (ns)    9.4       1.80      25.94
pH                              7.3 (a)     6.9 (b)   0.16       3.10
Turbidity (NTU)                 0.1 (b)    28.7 (a)   4.67      44.17
Hardness (mg Ca[CO.sub.3]/L)    0.2 (b)     0.3 (a)   0.05      27.70
Total Solids (mg/L)           228.8 (b)   488.1 (a)  43.27      16.51
HPC (log CFU/mL)                0.2 (b)     4.4 (a)   0.29      17.54
TCC (log CFU/mL)                0.1 (b)     3.4 (a)   0.20      15.98

ns--Means within a row are not significantly different

ab--Means within a row not followed by the same letter differ (P [less
than or equal to] 0.05)

(1)Fisher's Protected Least Significant Difference(P [less than or equal
to] 0.05)

(2)Coefficient of variation
Table 2

Mean of all the parameters of the discharge chiller water from two
catfish processing plants.

Parameter                          Water Source       LSD (1)  C.V. (2)
                                 Fresh    Recycled

Temp ([degrees]C)               7.7 (b)    10.1 (a)    1.34      20.62
pH                              7.1 (ns)    7.0        0.09       1.81
Turbidity (NTU)                16.6 (b)    33.7 (a)    6.40      34.84
Hardness (mg CaCO.sub.3]/L)     0.18 (b)    0.35 (a)   0.08      32.28
Settleable Solids (mg/L)        0.6 (ns)    0.0        0.66      290.84
Total Solids (mg/L)           990.3 (a)   503.1 (b)    173.9     31.84
HPC (log CFU/mL)                3.6 (b)     4.3 (a)    0.28       9.83
TCC (log CFU/mL)                2.6 (b)     3.4 (a)    0.50      22.65

ns--Means within a row are not significantly different

ab--Means within a row not followed by the same letter differ (P [less
than or equal to]0.05)

(1)Fisher's Protected Least Significant Difference (P [less than are
equal to] 0.05)

(2)Coefficient of variation
Table 3

Percent survival of identified Gram negative micro-organisms in chiller
water after 30 seconds of treatment with 3 ppm ozone.

MICROORGANISM                    # ISOLATED  PERCENT (%)

Aeromonas hydrophilia                2          8.33
Providencia alcalifaciens            2          8.33
Acinetobacter baumannii              4         16.67
Unidentified Gram Negative Rods     15         66.67
Total                               23        100.00
Table 4

Mean hetertrophic plate count (HPC) and total coliform count (TCC) loads
of recycled chiller water after treatment with high frequency ultrasound
(HFU) and ozone.

Treatment     Contact    Conc.       HPC            TCC
             Time (min)  (ppm)  (long CFU/mL)  (long CFU/mL)

HfU              0                   4.4 (ns)       2.8 (ns)
                 10                  4.3            2.8
                 15                  4.2            3.1
LSD (1)                              0.24           0.85

OZONE            0         0         4.4 (a)        3.3 (ns)
                 5       7.68        3.0 (b)        2.9
                 10      7.68        2.7 (b)        2.5
LSD (1)                              0.24           0.85

HFU + OZONE   10 + 5                 2.8 (ns)       2.5 (a)
              10 + 10                2.8            2.5 (a)
              15 + 5                 2.9            3.5 (b)
              15 + 10                2.9            2.6 (a)
LSD (1)                              0.24           0.85

Treatment    Listeria spp.
             presence (2)

HfU                +
                   +
                   +
LSD (1)

OZONE              +
                   -

LSD (1)

HFU + OZONE        -
                   -
                   -
                   -
LSD (1)

ns--Means within a row are not significantly different

ab--Means within a row followed by the same letter differ (p [less than
or equal] 0.05)

(1)Fisher's Protected Least Significant Difference (P [less than or
equal] 0.05)

(2)Listeria spp. presence: +present, -absent
Table 5

Effect of pulsed light on pH, turbidity, and bacterial load of recycled
catfish chiller water.

Parameter        Untreated    Treated
pH                 7.1 (a)    7.4 (b)
Turbidity (NTU)   20.3 (a)   10.0 (b)
HPC (log           3.91 (a)   0.0 (b)
CFU/mL)
TCC (log           3.51 (a)   0.0 (b)
CFU/mL)

ab--Means within a row not followed by the same letter differ (P [less
than or equal to] 0.05)

LITERATURE CITED

APHA. 1992. Compendium of methods for the microbiological examination of foods. American Public Health Association, Washington, D.C.

APHA. 1998. Standard methods for the examination of water and wastewater. American Public Health Association, Washington, D.C.

Barbosa-Canovas, G.V., U.R. Pothakamury, E. Palou, and B.G. Swanson. 1998. Nonthermal preservation of foods. Marcel Dekker, New York.

Burleson, G.R., T.M. Murray, and M. Pollard. 1975. Inactivation of viruses and bacteria by ozone, with and without sonication. Appl. Microbiol. 29:340-344.

Chamul, R.S. 2000. Use of continuous and pulsed high-frequency ultrasound to disinfect groundwater Ph.D. Dissertation. Mississippi State University, Mississippi State, MS.

Crum, L.K. 1992. Ultrasound: Seen but not heard. Chem. Eng. 10:37-41

Dowdy, S., and S. Wearden. 1985. Statistics for research. John Wiley & Sons, New York.

EPA. 1999. Multimedia environmental compliance guide for food processors. U.S. Environmental Protection Agency 305-B-99-005. Washington, D.C.

EPA. 2001. Current drinking water standards. http://www.epa.gov/ safewater/mcl.html

Federal Register. 1995. Current good manufacturing practices in manufacturing, packing, or holding human food. 21 CFR Part 110.37.

Federal Register. 2001. Final rule, Electric Power Research Institute, Agriculture and Food Technology Alliance. Ozone in gaseous and aqueous phase as an antimicrobial agent in food, including meat and poultry 66 FR 33829.

Fernandes, C.F, G. Flick,. J.L. Silva, and T.A. McCaskey. 1997. Influence of processing schemes on indicative bacteria and quality of fresh aquacultured catfish fillets. J. Food Prot. 60:54-58.

Figueroa, J.E., J.L. Silva, T. Kim, J. Boeger, and R. Cover. 2002. Use of pulsed-light to treat raw channel catfish fillets. J. Miss. Acad. Sci. 47:113-118.

Gray, N.F. 1994. Drinking water quality: Problems and solutions. John Wiley & Sons, New York.

Kim, T.J., J.L. Silva, R.S. Chamul, and T.C. Chen. 2000. Influence of ozone, hydrogen peroxide, or salt on microbial profile, TBARS and color of channel catfish fillets. J. Food Sci. 65:1210-1213.

Langlais, B., D.A. Reckhow, and D.R. Brink. 1990. Ozone in water treatment: Application and engineering. Lewis Publishers, Chelsea, Michigan.

MacGregor, S.J., N. J. Rowan, L. McIlvaney, J.G. Anderson, R.A. Fouracre, and O. Farish. 1998. Light inactivation of food-related pathogenic bacteria using a pulsed power source. Appl. Environ. Microbiol. 27:67-70.

Mason, T.J. 1993. Sonochemistry. A technology for tomorrow Chem. Ind. 2:47-50.

McGhee, T.J. 1991. Water supply and sewerage. McGraw-Hill, Inc. New York.

Nunez, A.L. 1995. Microbial loads in channel catfish processes and a HACCP program for raw fillets. M.S. Thesis. Mississippi State University, Mississippi State, MS.

Phull, S.S., A.P. Newman, J.P. Lorimer, B. Pollet, and T.J. Mason. 1997. The development and evaluation of ultrasound in the biocidal treatment of water. Ultrasonics Sonochem. 4:57-164.

PurePulse Technologies. 2002. http://www.purepulse.com. San Diego, CA.

Rowan, N. J., S.J. MacGregor, J.G. Anderson, R.A. Fouracre, L. McIlvaney, and O. Farish. 1999. Pulsed inactivation of food-related microorganisms. Appl. Environ. Microbiol. 65:1312-1315.

SAS. 1997. SAS User's guide basics. 5th ed. SAS Institute. Cary, NC.

Stephens, J.W. 1984. Design fundamentals for ozone absorption in bubble columns. M.S. Thesis. Mississippi State University. Mississippi State, MS.

TCI. 2001. The catfish institute. http://www.catfishinstitute.com

Watchalotone, W. 1996. Influence of process flow on microbial profile of channel catfish and of selected antimicrobials on its shelf life. M.S. Thesis. Mississippi State University, Mississippi State, MS.

WHO. 1985. Guidelines for the quality assessment of potable water. PAHO/OMS Vol. 1. http://www.who.int

Juan L. Silva (1)

(1.) Author for correspondence. E-mail: [e-mail]

Approved for publication as Journal Article No. 10107 of the Mississippi Agricultural and Forestry Experiment Station, Mississippi State University. This work was supported in part by USDA-ARS Grant No. 92-3850-7110 and by the Mississippi Agricultural and Forestry Experiment Station Project Number MIS-0889.