JOURNAL OF AGRICULTURAL RESEARCH AND EXTENSION Office of Agricultural Research & Extension Maejo University

Vol. 30 No.3 (Suppl.) June – September 2013 ISSN 0125-8850

Killng Varoa Mite by Grooming Behavior of Russian and Thai Honey Bees Boonmee Kavinseksan 1-13

Control of Off-flavor Cyanobacteria in Ponds using Nile Tilapia (Oreochromis niloticus) and Charcoal Bioreactor System Redel Gutierrez, Niwooti Whangchai, Khomsan Ruangrit and Tomoaki Itayama 14-28

Medicinally Potential of malabarica (L.) R. Br. in Comparison with a Porometer Rameshprabu Ramaraj and Yuwalee Unpaprom 29-39

Isolation and Identification of Cyanobacteria from a Freshwater Aquaculture Pond in Northern Thailand Dong Xia, Norio Iwami, Korntip Kannika, Chayarat Pleumsumran Sirapran Fakrajang, Chayaporn Teecharernwong, Redel Gutierrez Zhong Junsheng, Niwooti Whangchai and Tomoaki Itayama 40-48

Carbon Footprint of Central Canteen of Mahidol University Salaya Campus, Thailand Sayam Aroonsrimorakot, Chumporn Yuwaree, Chumlong Arunlertaree Rungjarus Hutajareorn and Tarinee Buadit 49-55

Mathematical Model of Freeze Drying on Mango Sakawduan Keawdam, Chanawat Nitatwichit, Jatupong Varith and Somkiat Jaturonglumlert 56-67

Fixed Deep Beds Drying of Black Pepper: A Comparative Study between a Normal Airflow and Reverse Airflow Phirunrat Thaisamak, Wipa Teppinta, Chanawat Nitatwichit Jatupong Varith and Somkiat Jaturonglumlert 68-79

Operation Cost Reduction for Industrial Pepper Powder Drying with Alternative Hot-air during Drying Process Wipa Teppinta, Jatupong Varith, Somkiat Jaturonglumlert Phirunrat Thaisamakand and Chanawat Nitatwichit 80-87 JOURNAL OF AGRICULTURAL RESEARCH AND EXTENSION

Honorable Consultants: Asst. Prof. Dr. Chamnian Yosraj Asst. Prof. Pawin Manochai Assoc. Prof. Dr. Duang Buddhasukh Dr. Waree Chaithep

Directing Editors: Assoc. Prof. Dr. Yongyut Khamsee Assoc. Prof. Jamnian Bunmark Asst. Prof. Dr. Jiraporn Intasan Dr. Sureewan Mekkamol

Editor-in Chief: Dr. Weerasak Prokati

Editorial Board: Prof. Dr. Tanongkiat Kiatsiriroat Prof. Dr. Siriwat Wongsiri Prof. Dr. Tomoaki Itayama Prof. Dr. Chen Li Hong Assoc. Prof. Dr. Nakao Nomura Assoc. Prof. Dr. Pramot Seetakoses Assoc. Prof. Dr. Duang Buddhasukh Assoc. Prof. Dr. Morakot Sukechotiratana Assoc. Prof. Dr. Sittisin Bovonsombut Assoc. Prof. Dr. Nopmanee Topoonyanont Assoc. Prof. Dr. Niwooti Wongchai Assoc. Prof. Dr. Weerachai Phutdhawong Asst. Prof. Dr. Jatuphong Varith Asst. Prof. Dr. Somkiat Jaturonglumlert Dr. Paisarn Kanchanawong Dr. Louise Label Dr. Yuwalee Anpapom Dr. Udomluk Sompong

Operation committee: Ms. Varee Rahong Mrs. Thanyarat Thawatmongkonsuk Ms. Rungsima Ampawan Mrs. Thipsuda Pookmanee Mr. Somyot Meesuk Ms. Ampar Sansai Mrs. Chiranan Senanan Ms. Dissawan Sutadsunti PR and Publishing: Mr. Prinya Painusa Mr. Prasit Chaikum Mrs. Prapaisri Thongjang Mrs. Suree Apichai

Journal of Agricultural Research and Extension is a publication of the Office of Agricultural Research and Extension, Maejo University and is intended to make available the results of technical work in the agricultural and related social sciences. Articles are contributed by MJU faculty members as well as by relevant the general public. Journal of Agricultural Research and Extension is published three times per year, contact with the Journal should be addressed to:

The Editor, Journal of Agricultural Research and Extension Innovation and Technology Transfer Section, Office of Agricultural Research and Extension Maejo University, Chiang Mai 50290, Thailand

Tel: +66-53-87-3932 Fax: +66-53-878-106 E-mail: [email protected] Website: www.rae.mju.ac.th Editor’s Note

The International Conference on Interdisciplinary Research and Development (ICIRD) in ASEAN Universities is hosted by Maejo University and The Interdisciplinary Network of the Royal Institute of Thailand under the Royal Patronage of HRH Princess Maha Chakri Sirindhorn, and is co-hosted by The University of Interdisciplinary Studies (UIDS), Chiang Mai University, North Chiang Mai University, Rajamangala University of Technology Isan, Southeast Asian Regional Center for graduate Study and Research in Agriculture (SEARCA), The Graduate School of the University of the Philippines Los Baños (UPLB), Mahidol University, and University of California Davis (UCD). This conference aims to make the gathering an event where scholars from different universities and educational institutions can exhibit, share and promote the advancement, experiences and insights from their research around the world leading to applications on humankind touched by these ASEAN universities. A stream of more than 300 academic articles and 400 participants around the globe has been entrusted into our collaboration. We hope this conference will stand for a profitable collaborative beginning and these scholarly papers will intrigue us and help stimulating discussions, thoughts and collaboration for our future research and development.

Siriwat Wongsiri Guest Editor Niwooti Whangchai and Somkiat Jaturonglumlert Associate Editors

Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Killing Varroa Mite by Grooming Behavior of Russian and Thai Honey Bees

Boonmee Kavinseksan Biology Program, Department of Science, Faculty of Science and Technology, Bansomdejchaopraya Rajabhat University, Bangkok, Thailand 10600 Corresponding author: [email protected]

Abstract

The purpose of this research was to the compare efficiency of killing Varroa mite according to grooming behavior of Russian and Thai honey bees. 20 Russian (Apis mellifera) and 20 Thai honey bee (domestic) colonies (A. mellifera) of Italian honey bee hybrids were used in this research. The experiment was conducted at an apiary in Chiang Mai, Thailand, on September 2012-January 2013. The number of dead Varroa mites in each colony was monitored every week using mite collecting boards with wire screens to prevent bees from carrying away debris. The boards were coated with a thin film of vegetable oil and maintained in the hives for periods of 7 days until the colonies died or to the end of the experiment (12 months). Dead mites on the boards were retrieved using a fine paint brush. They were examined and classified by type of injuries (using a stereomicroscope at 40X magnification). Injuries to the mites were classified as: (1) injured legs (missing legs or parts of legs), or (2) injured legs + injured body. The present results showed that the Russian bees have more efficient grooming behavior that killed more Varroa mites than the Thai bees. The average injured mite percentage of the Russian colonies (36.9±1.8%) was significantly higher (p≤0.05) than that of the Thai colonies (27.8±1.9%) (Mean±Standard error). For injury types, the average percentage of mites that had only leg injuries from the Russian (48.9±4.2%) and Thai (58.9±3.7%) colonies was not significantly different (p≤0.05). Mites that had both injured legs and body in the Russian and Thai debris were 51.11±4.2 and 41.1±3.7%, respectively (no significant difference p≤0.05).

Keywords: grooming behavior, Varroa mite, Apis mellifera, Russian honey bee, Thai honey bee

Introduction resulting in reduction of bee body weight and longevity (Figure 2). Wongsiri et al. (1989) reported Varroa jacobsoni and V. destructor are the that Varroa mites are a natural brood parasite of most economically disruption mite pests in A. cerana, a native honey bee species of Asia, beehives; causing serious damage to Apis but is not considered to be a severe pest of its mellifera beekeeping industries in the world. The natural host, A. cerana, due to the bee species mites feed on haemolymph of bees (Figure 1); defensive mechanisms to this mite such as 1 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13 absconding, hygienic behavior and grooming of chemical and biotechnical methods provide behavior. some relief to colonies but not offers complete Grooming behavior of honey bees (Apis eradication of mites (Wongsiri et al., 1987; spp.) is a natural defense against parasitic mites Tangkanasing et al., 1988). In addition, these (Rath and Delfinado-Bake, 1990; Koeniger et al., methods are either labor intensive, costly, 2002). This behavior by A. dorsata and A. cerana reducing bee populations or contaminating bee against Tropilaelaps clareae and Varroa mites has products. been reported by several researchers (Koeniger The use of A. mellifera strains resistant to et al., 1980; Wongsiri et al., 1987; Rath, 1991; Varroa mites have been thought to be a better Kavinseksan, 2003; Kavinseksan, 2011). A. dorsata solution to the Varroa problem. The benefits of has efficient grooming behavior which removes using resistant A. mellifera against parasitic mites mites from their bodies and kills them, indicated include: less chance of contaminating bee products by the numbers of injured Varroa mites (37.5%) with undesirable chemicals, low cost of labor and T. koenigerum (76%) and T. clareae (73-93%) in materials and less risk of the mite developing debris from A. dorsata colonies (Rath and Delfinado- resistance to acaricides (Kavinseksan et al., 2006; Bake, 1990; Koeniger et al., 2002; Kavinseksan, Lensky et al., 2001; Kavinseksan et al., 2003; 2003; Kavinseksan, 2011; Kavinseksan, 2005) and Kavinseksan et al., 2004). However, no studies the low levels of infested adult bees (0.2%) in Thailand have been done to find A. mellifera (Kavinseksan et al., 2006). strains resistant to Varroa mites. The Russian When A. mellifera was introduced to the honey bee (A. mellifera) is known to be resistant to areas of A. cerana, Varroa mite successfully V. jacobsoni, V. destructor, Acarapis woodi and switched from indigenous A. cerana host to T. clareae (Kavinseksan, 2003, Rinderer et al., 1997; A. mellifera and has become a serious problem to Rinderer et al., 1999; Rinderrer et al., 2000; Rinderrer A. mellifera colonies due to their lack of defense et al., 2001; Guzman et al., 1996; Guzman et al., mechanisms necessary to regulate mite populations 2001; Kavinseksan, 2012). However, resistance of in its colonies (Wongsiri et al., 1989). A. cerana the Russian bee to Varroa mites in Thailand has was distributed throughout Thailand. Thus, yet to be established. Therefore, this research was A. mellifera colonies in all regions could not avoid conducted to compare the efficiency of killing Varroa infestation by Varroa mites. Several studies have mite through grooming behavior of Russian and been conducted to find the effective control Thai honey bees. against Varroa mites in A. mellifera colonies. Chemical, physical, biotechnical and combinations

2 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Figure 1 Varroa mites feed on haemolymph of A. mellifera larva, pupa and adult bees.

Figure 2 Deformation of A. mellifera worker bees’ wings; resulting from Varroa mites feeding on haemolymph in larva and pupa stages.

Materials and Methods USA, while the Thai queen bees were provided by Supa’s apiary in Chiang Mai, Thailand. All queens The grooming behavior efficiency against of both strains were introduced into standard Varroa mites was compared between 20 Russian Langstroth hives comprised of about 15,000 adult (A. mellifera) and 20 Thai domestic (Italian honey worker bees. All colonies were provisioned with bee hybrids) colonies (Figure 3-4) in Chiang Mai, four frames of honey and and four empty Thailand, on September 2012-January 2013. The combs. Experimental procedures were carried out Russian queen bees were obtained from the about five months after queen introduction, insuring USDA-ARS Honey Bee Breeding, Genetics and that the grooming behavior was measured from the Physiology Laboratory in Baton Rouge, Louisiana, progeny of the experimental queens.

3 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Grooming behavior against Varroa mites of the Russian and Thai bees was evaluated on percentages of injured mites in hive debris. The number of dead Varroa mites in each colony were monitored weekly using mite collecting boards with wire screens to prevent bees from carrying away debris (Figure 5). The boards were coated with a thin film of vegetable oil and maintained in the hives for periods of 7 days until the colonies Figure 3 Russian honey bees died or to the end of the experiment (12 months). Dead mites on the boards were retrieved using a fine paint brush. They were examined and classified based on types of injuries (using a stereomicroscope at 40X magnification). Injuries to the mites were classified as: (1) injured legs (missing legs or parts of legs), or (2) injured legs +body. Data from each week of each bee strain were pooled together for final analysis. The data of injured mite percentages and injury types of the Figure 4 Thai honey bees mites in the debris were analyzed using t-tests (SPSS statistic program). The bee colonies were considered as the replications. The Pearson correlation coefficient was used to evaluate the relationships between the dead mite numbers and injured mite numbers in the debris of each bee strain.

Figure 5 A mite collecting board 4 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Results and Discussion colonies (37.5%) from the previous report by Kavinseksan (Kavinseksan, 2011). These suggest The results of this study showed that a total that the Russian honey bee had an effective number of Varroa mites were collected on bottom grooming behavior which destroys Varroa mites in board traps from 20 Russian and 20 Thai colonies similar levels with A. dorsata. were 2,693 and 3,832, respectively. The average The growth and development of Varroa number of dead mites in the Russian and Thai mites in A. mellifera colonies depend on several debris has no significant different (t=1.276, df= factors such as non-reproduction, hygienic and 38, Sig=0.210) with means of 134.7±15.2 and grooming behaviors of the honey bee hosts 191.6±41.9 mites per colony (Mean±Standard error), (Wongsiri et al., 1989; Rath and Delfinado-Bake, respectively (Table 1). The number of injured mites 1990; Kavinseksan, 2005; Kaviseksan et al., 2006; in the Russian and Thai debris did not significantly Kaviseksan et al., 2003; Kaviseksan et al., 2004; differ (t=0.325, df=38, Sig=0.747) with means of Kaviseksan et al., 2012; Buchler and Drescher, 49.7±5.0 and 46.4±8.7 mites per colony, respectively 1990; Woyke, 1990; Woyke et al., 2000; Woyke This is the first report on efficient grooming et al., 2004; Anderson, 1994; Spivak and Reuter, behavior of the Russian and Thai honey bees 1994; Kavinseksan, 2007; Kavinseksan, 2008; against Varroa mites in Thailand. The result Kavinseksan, 2011) Grooming behaviors of A. cerana showed that the Russian bees exhibited more and A. dorsata are effective to removing and kill mites efficient grooming behavior to kill Varroa mites from their bodies (Wongsiri, 1989; Kavinseksan, than that of the Thai bees (Figure 6), which is 2011). Self-cleaning (auto-grooming) behavior of indicated by the significant higher percentage A. dorsata in response to T. clareae infestation has (t=3.305, df=38, Sig=0.002) of injured mites in the been reported by Buchler et al. (1992). A. cerana debris from the Russian colonies than the Thai has a nest-mate-cleaning behavior (grooming colonies with means of 36.9±1.8 and 27.8±1.9%, behavior) between worker bees which is an respectively (Table 1). The amount of injured mites essential defense mechanism to prevent Varroa showed a similar pattern with the amount of the and Tropilaelaps populations in their colonies dead mites. The total injured Varroa mite percentage from reaching dangerous level (Buchler et al., 1992). in the Russian and Thai debris (27.8-36.9%) in Grooming behavior of A. cerana was applied for this study was lower than the percentage of injured biological control to regulate Varroa and Tropilaelaps Tropilaelaps mites in the debris from the Russian populations in infested A. mellifera colonies by and Thai colonies (71.2-77.1%) and in A. dorsata insertion of A. cerana pupae into A. mellifera colonies. debris (94.7%) compared to thefindings by Kavin The inserted pupae developed and emerged from seksan (2003, 2012). The percentage of injured the cells to become nurse bees (1-14 days-old). Varroa mites in the debris from the Russian They performed grooming behavior to remove and colonies (36.9%) was similar to that of A. dorsata kill Varroa and Tropilaelaps mites from bodies of

5 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

A. mellifera adult bees. Self-cleaning (auto-grooming) the Russian and Thai colonies have no significantly behavior by A. cerana in response to Varroa and different (t=1.803, df=38, Sig=0.079) with means Tropilaelaps infestation has also been reported of 48.9±4.2 and 58.9±3.7%, respectively (Table 1). (Wongsiri et al., 1987). The percentage of severely damaged legs and body In this study, Pearson correlation coefficient of Varroa mites in the debris from the Russian and showed significant higher levels of relationship Thai colonies was not significantly different (t=1.803, between the number of dead mite and the number df=38, Sig=0.079) with means of 51.1±4.2 and 41.1 of injured mites in the debris from the Russian ±3.7%, respectively (Figure 7). The percentage of (r=0.882) and Thai (r=0.986) colonies (Sig=0.000 for severely damaged Varroa mites (41.1-51.1%) in each strain). These results indicated that almost this study was higher than that of Tropilaelaps mites Varroa mites (88.2-98.6%) on the collecting boards in the Russian and Thai debris (17.1-20.6%) from of the Russian and Thai colonies were removed the previous report by Kavinseksan (2012). These from the bees bodies by their grooming behavior. suggest that the Russian and Thai bees had For injury types, the results showed that higher grooming behavior efficiency in destroying the percentage of dead mites with injured legs only, Varroa mites than Tropilaelaps mites. by missing legs or parts of legs in the debris of

6 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Table 1 Numbers and percentages of dead and injured Varroa mites in debris of Russian and Thai honey bee colonies (Mean±Standard error)

Injured types Bee No. of Total mites Injured mites / colony strains colonies Legs Legs+body

Numbers Percentages Numbers Percentages Numbers Percentages 1 193 59 30.6 24 40.7 35 59.3 2 132 30 22.7 6 20.0 24 80 3 108 48 44.4 30 62.5 18 37.5 4 98 36 36.7 20 55.6 16 44.4 5 138 68 34.8 40 58.8 28 41.2 6 96 30 31.3 18 60.0 12 40 7 96 36 37.5 24 66.7 12 33.3 8 100 50 50 10 20.0 40 80 9 110 60 54.5 20 33.3 40 66.7 10 108 36 33.3 20 55.6 16 44.4 Russian 11 80 32 40 8 25.0 24 75

12 72 24 33.3 18 75.0 6 25 13 386 118 30.6 48 40.7 70 59.3

14 88 20 22.7 4 20.0 16 80 15 162 72 44.4 45 62.5 27 37.5 16 196 72 36.7 40 55.6 32 44.4

17 138 48 34.8 20 41.7 28 58.3 18 128 40 31.3 24 60.0 16 40

19 144 54 37.5 45 83.3 9 16.7 20 120 60 50 24 40.0 36 60 Total 2,693 993 488 505 Average 134.7±15.2a 49.7±5.0a 36.9±1.8a 48.9±4.2a 51.1±4.2a 1 102 18 17.6 12 66.7 6 33.3 2 101 19 18.8 7 36.8 12 63.2 3 148 40 27 19 47.5 21 52.5 4 126 30 35.7 18 60.0 12 40.0 5 119 31 26.1 12 38.7 19 61.3 6 713 150 21 83 55.3 67 44.7 7 132 36 27.3 30 83.3 6 16.7 8 99 27 27.3 18 66.7 9 33.3 Thai 9 102 30 29.4 12 40.0 18 60.0 10 87 21 24.1 13 61.9 8 38.1 11 88 32 36.4 28 87.5 4 12.5 12 84 32 38.1 20 62.5 12 37.5 13 68 12 17.6 10 83.3 2 16.7 14 202 38 18.8 14 36.8 24 63.2 15 296 80 54 38 47.5 42 52.5 16 84 30 35.7 18 60.0 12 40.0 17 238 62 26.1 24 38.7 38 61.3 18 713 150 21 83 55.3 67 44.7 19 132 36 27.3 30 83.3 6 16.7 20 198 54 27.3 36 66.7 18 33.3 Total 3,832 928 525 403 Average 191.6±41.9a 46.4±8.7a 27.8±1.9b 58.9±3.7a 41.1±3.7a

Similar letters in the same column are not significantly different at the 0.05 level.

7 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

to be the most important limiting factors to the development and expansion of A. mellifera beekeeping in Thailand and tropical Asia (De Jong et al., 1982; Nyein and Zmarlicki, 1982; Woyke, 1985). However, several researchers suggest that T. clareae may be more destructive than Varroa mites to A. mellifera colonies because the numbers of T. clareae are often higher than those of Varroa mites in Thailand (Wongsiri et al., 1989; Burgett Figure 6 Nest-mate-grooming behavior between et al., 1983). Thai worker bees (A. mellifera). In Thailand, A. mellifera was introduced for the first time in the early 1940s and again in 1953 but maintainance was not successful until the early 1970s (Wongsiri and Chen, 1995). Therefore, A. mellifera in Thailand (Thai honey bee) and Varroa mites have coexisted and coevolved for about 40 years. This time period might be too short for the Thai honey bees to Figure 7 Injured Varroa mites in debris of Russian develop a high level of genetic resistance or and Thai honey bee colonies. natural defensive mechanisms to regulate Varroa populations in their colonies. In the case of the Since the western honey bee A. mellifera Russian honey bee, which has coexisted with has been introduced to the areas with A. cerana, V. jacobsoni for more than 160 years has developed Varroa mites successfully switched from its a high level of genetic resistance and natural natural host, and has become a dangerous pest defensive mechanisms to regulate Varroa populations of A. melliera (Wongsiri et al., 1987; Rath, 1991). in their colonies, ie grooming behavior. In the USA, Also, when A. melliera was introduced to tropical the Russian honey bees have been used for Asia, T. clareae successfully switched hosts from commercial beekeeping because this honey bee A. dorsata to A. melliera and has become a severe strain has genetic resistance and natural defensive pest of A. mellifera (Rath et al., 1990). Wongsiri mechanisms to V. jacobsoni, V. destructor and et al. (1989) reported that A. mellifera lacks essential A. woodi (Rinderrer et al., 1997; Rinderrer et al., defensive mechanisms to limit the population growth 1999; Rinderrer et al., 2000; Rinderrer et al., of Varroa mites and T. clareae in its colony. Both 2001; Guzman et al., 1996; Guzman et al., 2001; Varroa and Tropilaelaps mites have been considered Danka et al., 1995).

8 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Grooming behavior of honey bees is kill them than that of the Thai bee. This is regulated by genes; including hygienic behavior indicated by the percentage of injured mites in and non-reproduction of mites (Rothenbuhler, 1964; the debris from the Russian (36.9±1.8%) and Thai Harbo and Harris, 1999; Harris and Harbo, 2001). (27.8±1.9%) colonies. For injury types, the dead Furthermore, most strains of A. mellifera have no mites in the Russian and Thai debris showed the exhibit strong grooming behavior and artificial similar type of damages. Percentage of dead selection is required to enhance the trait. Both mites with only injured legs (48.9-58.9%), and strains in this study have high levels of grooming percentage of severely damaged mites on both behavior and this resistance has risen as a result legs and body (41.1-51.1%) was not significantly of being maintained and propagated in areas with different (p≤0.05). The Pearson correlation parasitic mites, although neither strain has been coefficient between the number of dead and specifically selected for grooming behavior. injured mites in the debris of the Russian and This characteristic may have resulted from Thai Thai colonies were 0.882 and 0.986, respectively beekeepers propagation from their strongest (Sig=0.000 for each strain), or 88.2-98.6% of colonies in areas infested by Varroa mites. Varroa mites were removed from the bee bodies Beekeepers may have inadvertently selected for by grooming behavior of the Russian and Thai strong grooming behavior and perhaps other bees. mechanisms resistance to Varroa mites such as hygienic behavior and non-reproduction of mites. Recommendations Most importantly, this study indicates that both for Russian and Thai strains of A. mellifera Based on the success of the Russian possess grooming traits, which may be helpful in bees at controlling Varroa mites by their grooming regulating Varroa mite population in their colonies. behavior, beekeepers should use the Russian A selection program with the commercial Russian bees for commercial A. mellifera beekeeping to and Thai honey bee strains specifically directed solve the Varroa infestation problem in A. mellifera toward increasing their resistance to Varroa mites colonies. The use of artificial selection for resistant may produce stocks with sufficient resistance to stocks and breeding programs (natural or artificial slow the growth of Varroa mite populations, which breeding) to increase the level of genetic resistance would reduce levels of acaricide use. in A. mellifera is also recommended. For selections, A. mellifera that has coexisted with Varroa mites Conclusion for the longest time should be tested for resistance to the mite. A. mellifera may have genetic traites Results presented in this study show that which impart resistance to Varroa mites that have the Russian bee has significantly more efficient already been selected to some degree in such grooming behavior to remove Varroa mites and populations. After that, breeding program should be

9 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13 used to produce resistant hybrids to increase Danka, R.G., T.E. Rinderer, V.N. Kuznetsov and the level of genetic resistance in A. mellifera to G.T. Delatte. 1995. A USDA-ARS Varroa mites. project to evaluate resistance to Varroa jacobsoni by honey bees of Far-eastern Acknowledgements Russia. Am. Bee J. 135(11): 746-748. De Guzman, L.I., T.E. Rinderer, G.T. Delatte and I am grateful to National Research Council R.E. Macchiavelli. 1996. Varroa of Thailand (NRCT) for financial support. Also, jacobsoni Oudemans tolerance in appreciation is directed to Professor Dr. Siriwat selected stocks of Apis mellifera L. Wongsiri, Professor Dr. Thomas E. Rinderer, and J. Apidologie. 27: 193-210. Dr. Lilia De Guzman for their suggestions on De Guzman, L.I., T.E. Rinderer, J.A. Stelzer, research techniques. L. Beaman, G. Delatte and C. Harper. 2001. Hygienic behavior by honey bees References from Far-eastern Russia. Am. Bee J. 141: 58-60. Anderson, D.L. 1994. Non-reproduction of Harbo, J.R. and J.W. Harris. 1999. Heritability in Varroa jacobsoni in Apis mellifera colonies honey bees (Hymenoptera: Apidae) of in Papua New Guinea and Indonesia. characteristics associated with resistance J. Apidologie. 25: 412- 421. to Varroa jacobsoni (Mesostigmata: Burgett, D.M., P. Akratanakul and R.A. Morse Varroidae). J. Econ. Entomol. 1983. Tropilaelaps clareae: a parasite of 92(2): 261-265. honeybees in southeast Asia. Bee Wld. Harris, J.W. and J.R. Harbo. 2001. Natural & 64: 25-28. suppressed reproduction of Varroa. Buchler, R. and W. Drescher. 1990. Variance Bee Culture. 34-38. and heritability of the capped Jong, D. De, R.A. Morse and G.C. Eickwort. developmental stage in European Apis 1982. Mite pests of honey bees. mellifera L. colonies and its correlation J. Annu. Rev. Entomol. 27: 229-252. with increased Varroa jacobsoni Oud. Kavinseksan, B. 2003. Defense mechanisms Infestation. J. Apic. Res. 29: 172-176. of Apis dorsata Fabricius and ARS Buchler, R., W. Drescher and I. Tornier. 1992. Primorsky honey bee Apis mellifera Grooming behavior of Apis cerana, Apis Linnaeus to the bee mite Tropilaelaps mellifera and Apis dorsata and its effect clareae Delfinado and Baker. PhD. on the parasitic mites Varroa jacobsoni Thesis. Chulalongkorn University. and Tropilaelaps clareae. J. Exp. Appl. Acarol. 16: 313-319.

10 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Kavinseksan, B., S. Wongsiri, L. De Guzman and Kavinseksan, B. 2011. Defense mechanisms T.E. Rinderer. 2003. Absence of Apis dorsata Fabricius to Varroa mites of Tropilaelaps infestation from recent and diseases. J. of Apiculture. swarms of Apis dorsata in Thailand. 26(1): 15-19. J. Apic. Res. 49-50. Kavinseksan, B. 2011. Resistance of Russian Kavinseksan, B., S. Wongsiri, T.E. Rinderer and and Thai honey bees to Tropilaelaps L. De Guzman. 2004. Comparison of clareae (in Thai). Advanced Science hygienic behavior of Thai commercial and Journal. 11(2): 67- 81. ARS Russian honey bees. Am. Bee J. Kavinseksan, B. 2012. Grooming behavior of 144(11): 870-872. ARS Russian and Thai domestic honey Kavinseksan, B. 2005. Defense mechanisms bees against Tropilaelaps clareae and its of Apis dorsata Fabricius to the bee mite sex ratio. Journal of Apiculture. Tropilaelaps koenigerum (in Thai). 27(1): 1-7. Advanced Science Journal. 5(2): 67-83. Koeniger, N. and G. Koeniger. 1980. Kavinseksan, B., S. Wongsiri and T.E. Rinderer. Observations and experiments on 2006. Tropilaelaps clareae populations migration and dance communication of in new, established and deserted nests Apis dorsata in . J. Apic. Res. of Apis dorsata in Thailand (in Thai). 19(1): 21-34. Advanced Science Journal. 6(1): 52-65. Koeniger, G., N. Koeniger, D.L. Anderson, Kavinseksan, B. 2007. Non-reproduction and C. Lekprayon and S. Tingek. 2002. reproduction of Tropilaelaps clareae Mites from debris and sealed brood cells Delfinado and Baker in Apis dorsata of Apis dorsata colonies in Sabah Fabricius (in Thai), pp. 205-212. In (Borneo) , including a new The National Meeting on Teaching and haplotype of Varroa jacobsoni. Learning of Science using Integrated J. Apidologie. 33: 15-24. Approaches: From Local Science to Lensky, Y., U. Dechmani, U. Jirasavetakul and Learning Centers. Chingrai Rajabhat S. Tiewtrakul. 2001. The effect University, Thailand. of fluvalinate on Varroa jacobsoni (Oud.) Kavinseksan, B. 2008. A Varroa mite defense and Tropilaelaps clareae (Delfinado mechanism and the hygienic behavior and Baker) in Apis mellifera colonies of Apis dorsata Fabricius (in Thai). in north Thailand. pp. 176-182. In Advanced Science Journal. Proceedings of the third asian 8(2): 125-137. apicultural association conference on bee research and beekeeping development. Hanoi.

11 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Nyein, M.M. and C. Zmarlicki. 1982. Control of Rinderer, T.E. 2001. Multi-state field trials of mites in European bees in Burma. ARS Russian honey bees: 1. Responses Am. Bee J. 122: 638-639. to Varroa destructor 1999, 2000. Rath, W. 1981. Investigation on the parasite Am. Bee J. 141(9): 658-661. mites Varroa jacobsoni Oud. and Rothenbuhler, W.C. 1964. Behavior genetics Tropilaelaps clareae Delfinado and of nest cleaning in honey bees:

Baker and their hosts Apis cerana IV. Responses of F1 and backcross Fabr., Apis dorsata Fabr. and Apis generations to disease-killed brood. mellifera L. PhD. Thesis. Matematisch Am. Zoologist. 4: 111-123. Naturwissen, Fakulat der Rheinsischen Spivak, M. and G.S. Reuter. 1998. Honey bee Friedrich Wilhelms Universitat. hygienic behavior. Am. Bee J. Rath, W. and M. Delfinado-Bake. 1990. Analysis 138(4): 283-286. of Tropilaelaps clareae populations from Tangkanasing P., S. Wongsiri and the debris of Apis dorsata and Apis S. Vongsamanod. 1988. Integrated mellifera in Thailand. Proceedings of the control of Varroa jacobsoni and apimondia symposium recent research Tropilaelaps clareae in bee hives in on bee pathology. Gent. Thailand. pp. 409-412. In Africanized Rinderer, T.E., V. N. Kuznetsov, R.G. Danka Honey Bee and Bee Mites (Ed. G. R. and G.T. Detatte. 1997. An importation Needham, R.E. Page, M. Delfinado-Baker of potentially Varroa-resistant honey bees and C.E. Bowman). New York: from Far-Eastern Russia. Am. Bee J. John Wiley & Sons. 137(11): 787-789. Wongsiri, S., P. Tangkanasing and S. Vongsamanodes. Rinderer, T.E., G.T. Delatte, L.I. De Guzman, 1987. Effectiveness of Asuntol® J. Williams, J. Stelzer and V.N. Kuznetsov. (coumaphos), Perizin® (coumaphos), 1999. Evaluations of the Varroa- Mitac® (amitraz) and powder of sulphur resistance of the honey bees imported with naphthalene for the control of bee from Far-Eastern Russia. mites (Varroa jacobsoni and Tropilaelaps Am. Bee J. 139(4): 287-290. clareae) in Thailand. Proceedings of the Rinderer, T.E., L.I. De Guzman, J. Harris, XXXI st International Apicultural V. Kuznetsov, G.T. Delatte, J.A. Stelzer Congress of Apicultural Congress of and Beaman L. 2000. The release of APIMONDIA. Warsaw. ARS Russian honey bees. Am. Bee J. 140(4): 305-307.

12 Journal of Agr. Research & Extension 30(3) (Suppl.): 1-13

Wongsiri, S., P. Tangkanasing and H.A. Sylvester. Woyke, J., J. Wilde and M. Wilde. 2000. 1987. Mites, pests and beekeeping with Swarming, migrating and absconding of Apis cerana and Apis mellifera in Apis dorsata colonies. Proceedings of Thailand. J. Am. Bee. 127(7): 500-503 the seventh international conference on Wongsiri, S., P. Tangkanasing and H.A. Sylvester. tropical bees: management and 1989. The resistance behavior of Apis diversity and fifth asian apicultural cerana against Tropilaelaps clareae. association conference. Proceedings of the First Asia Pacific Woyke, J. 1985. Tropilaelaps clareae, a serious Conference of Entomology. pest of Apis mellifera in the tropics but Chaing Mai. pp. 828-836. not dangerous for apiculture in temperate Wongsiri, S. and P. Chen. 1995. Effects of zones. Am. Bee J. 125(7): 497-499. agricultural development on honey bees Woyke J., J. Wilde and C.C. Reddy. 2004. in Thailand. Bee Wld. 76(1): 3-5. Open-air-nesting honey bees Apis dorsata Woyke, J. 1990. Biology and control of the and Apis laboriosa differ from the cavity- parasitic bee mite Tropilaelaps clareae. nesting Apis mellifera and Apis cerana in Proceedings of the apimondia brood hygiene behaviour. Journal of symposium recent research on bee Invertebrate Pathology. 86: 1-6. pathology. Gent.

13 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Control of Off-flavor Cyanobacteria in Ponds using Nile tilapia (Oreochromis niloticus) and Charcoal Bioreactor System

Redel Gutierrez 1,2*, Niwooti Whangchai 1, Khomsan Ruangrit 3 and Tomoaki Itayama4 1Faculty of Fisheries Technology and Aquatic Resources, Maejo University, Chiang Mai, Thailand 50290 2College of Arts and Sciences, Central Luzon State University, Science City of Munoz, Nueva Ecija, Philippines 3120 3Department of Biology, Faculty of Science, Chiang Mai University, Chiang Mai, Thailand 50200 4Graduate School of Engineering, Nagasaki University, Nagasaki, Japan 852-8521 *Corresponding author: [email protected]

Abstract

Off-flavors of cyanobacterial origins represent one of the most significant economic problems in the aquaculture industry especially in catfish production. This study investigated the combined effects of Nile tilapia (Oreochromus niloticus) and charcoal bioreactor in controlling cyanobacteria and reducing off- flavor in Thai hybrid pangasius (Pangasinodon gigas x Pangasinodon hypothalmus) ponds. The experiments were conducted in twelve 100-m2 earthen ponds, eight of which were integrated with tilapia in cages (2 units) equipped with a charcoal bioreactor system each. Nile tilapia (0, 0.5 and 1.0 m -2 density) and Thai hybrid pangasius (3.0 m-2 density) were subjected to a no-feeding and non-restricted feeding regime, respectively for 5 months. Results show that both Thai hybrid pangasius and Nile tilapia grew steadily in all treatments over the 5-month culture cycle. No significant differences were observed in the growth performance of Thai hybrid pangasius among all treatments (p>0.05). Similarly, Nile tilapia growth was not significantly different between high (T3-1.0 fish m-2) and low (T2-0.5 fish m-2) density treatments (p>0.05). Average weight gained by Thai hybrid panga ranged from 389.3–442.1 g whist Nile tilapia gained 49.00 g and 52.25 g for T2 and T3, respectively, indicating growth possibly from effective grazing on algae and cyanobacteria in water. Cyanobacterial cell density in integrated systems with bioreactors decreased in the following order: pond water > cage water > bioreactor effluent. Total off-flavor levels (GSM + MIB) in pond waters of T2 and T3 were significantly lower (p<0.05) than T1 (control). However, off-flavor levels and water quality were comparable between low (T2) and high (T3) tilapia density treatments. This study suggests that the combined intervention applied in Thai hybrid panga ponds is practically effective in controlling cyanobacteria and reducing off- flavor in water.

Keywords: Nile tilapia, Thai hybrid panga, charcoal bioreactor, off-flavor, cyanobacteria

14 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Introduction Xie and Liu, 2001; Jin et al., 2006; Xie, 2003). This strategy is common in tropical regions, where Intensive-culture ponds, such as Thai panga the zooplankton does not seem able to control catfish (Pangasiusnodon hypophthalmus x Pangasius algal biomass, a role that is played instead mainly boucorti) production ponds, are abundant in rich by omnivorous, filter-feeding fishes (Komarkova, nutrients and planktonic organisms. Phytoplankton 1998). However, research on the use of silver often increase excessively in these eutrophic water carp (Hypophthalmichthys molitrix) as filter feeding bodies, causing reduced water transparency and fish has revealed several disadvantages. Silver the production and release of toxins and off-flavor carp reduces zooplankton and large algae but compounds as secondary metabolites. Off-flavor enhances micro algal growth, which in turn problems caused by phytoplankton particularly increases algal biomass rather than reducing it. cyanobacteria (and actinomycetes) in Thai Panga Moreover, digestion efficiency on cyanobacterial farms are seasonal occurrences which reduce colonies such as Microcystis is low (25-30%) market acceptability of Thai panga and can delay (Chen et al., 1990). The metabolic activity of harvesting resulting in a loss millions of bath per phytoplankton after gut passage in silver carp and year. One way to control excessive phytoplankton bighead carp remains unaffected or even increases abundance is to introduce filter feeding fishes (Miura et al., 1985; Friedland et al., 2005; Kolmakov (planktivorous or omnivorous) cultured in the same and Gladyshev, 2003; Gavel et al., 2004). On pond along with Thai panga catfish. This scheme the other hand, tilapia can be more effectively used could mitigate environmental impact of intensive for algal control due to their specific feeding and aquaculture and to increase the economic incentive digestion capacities (Lu et al., 2006). Unlike silver of integrated culture as well as reduce the nuisance carp and bighead, tilapia is omnivorous (Bowen, cyanobacteria to minimize off-flavor in cultured 1983; Chapman and Fernando, 1994) and capable fish. Also, reduced phytoplankton abundance will of filtering, particle capturing, and scraping. lead to an improvement of water quality and clearer These characters enable tilapia a broad feeding water. habit and feeding capacity. Nile tilapia, Oreochromis Biomanipulation is the term applied to niloticus, is known to have an extremely low pH such manipulations of the biota and their habitats value (2.5 to 1.0) in its stomach (Getachew, 1989) to facilitate biological interactions that result in therefore enhanced damage of ingested cyanobacteria the reduction of excessive algal biomass–in is probable. More than 95% loss of viability of particular, of cyanobacteria (Shapiro, 1990; Carpenter Microcystis colonies after gut passage has been and Kitchell, 1992). Based on “top-down control”, recently demonstrated (Jančula et al., 2008). Lu an alternative form of biomanipulation is to use filter et al. (2006) also demonstrated that tilapia had feeding fish rather than herbivorous zooplankton, high ingestion rates and digestion efficiencies to reduce phytoplankton biomass (Starling, 1993; for Microcystis, which is in agreement with several

15 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28 previous findings (Bowen, 1983; Moriarty, 1973), the bioreactor used by Bi et al. (2013) with further demonstrating that tilapia is a better fish modifications, was employed in tilapia cages than other filter feeders for algal control in natural integrated in Thai panga ponds. freshwater waters. The purpose of this study was to evaluate Biomanipulation can also be used in the feasibility of reducing cyanobacteria and combination with other bio-physical methods of the incidence of fish off-flavors in Thai panga controlling or removing cyanobacteria in ponds. ponds by manipulating pond phytoplankton biomass Itayama et al. (2008) developed a method for and taxonomic composition using a combination cyanobacterial control involving a bioreactor of biological and physical control measures. inhabited by microfauna such as the micro flagerata Monas guttula and the oligochaete Aeolosoma Materials and Methods hemprichi. The bioreactor effectively suppressed the proliferation of cyanobacteria in an experimental Experimental set-up pond by the continuous removal of cyanobacterial The study was carried out for a 5-month cells from the pond water. The major targets of period at Thai Panga Farm (TPF) in Kalasin the bioreactor system were small reservoirs and Province, northern Thailand. Twelve experimental recreation ponds since large lakes and reservoirs green water ponds, each with a 100-m2 area and require large scale bioreactor systems which are 2.0 m depth, were used (Table 1). Three treatments very expensive (Itayama et al., 2008). Another with 4 replicates in a completely randomized design system, referred to as “Bio-fence” was designed were performed as follows: T1–Thai panga catfish and proposed by a Japanese company to perform in ponds at 20,000 fish ha-1. T2-Thai panga catfish clarification of a partial surface area of large lakes in ponds at 20,000 fish ha-1 integrated with Nile and reservoirs (Itayama et al., 2008). A water zone tilapia in cages at 3500 fish ha-1 and; T3–Thai panga is surrounded with the biofence consisted with catfish in ponds at 20,000 fish ha-1 integrated with bio-carriers, which are habitats of predators such as Nile tilapia in cages at 7000 fish ha-1. Fish stocked A. hemplichi for cyanobacteria. The cyanobacteria (young adult Nile tilapia and Thai panga, with overall from the contaminated water can be removed average initial weights of 204.7 g and 290.0 g, by pumping up water through the bio-carriers of respectively) in each pond and cage were counted the bio-fence. Bi et al. (2013) pilot-tested a bioreactor and weighed at the beginning and at the end of based on the biofence model using charcoal carriers the experiment, when ponds were drained and all to decompose cyanobacteria and cyanotoxins in fish collected. Nile tilapia received no food during an aquaculture pond in northern Thailand and the experiment except for plankton naturally available found it effective. In this study, a similar design of inside the cage whilst Thai panga were subjected

16 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Table 1 Pond properties and set-up Treatments

T1 T2 T3 Pond area (m2) 100 100 100 Cage dimension (m3) - 2.5 × 2 × 2 2.5 × 2 × 2 No. of tilapia stocked 0 50 100 No. of Thai panga stocked 300 300 300 Tilapia average weight (g) - 203.50 206.75 Thai panga average weight (g) 230.15 212.12 212.75 With charcoal bioreactor No Yes Yes to a non-restricted feeding regime. Basket-type board with its inner circumference fitting the smaller wood charcoal bioreactor was installed inside basket and its perimeter fitting onto the sidewall of each tilapia cage. Each pond in T2 and T3 had 2 the main basket. To prevent the basket bioreactor cages situated in the middle separated 2.5 m to tumble in water and maintain its balance due to away from each other. its lightness, sufficient amount of pebble stones were placed on top the plastic board to forcibly Description of the charcoal bioreactor system immerse the basket in the water. Four 2.5-L A fabricated charcoal basket bioreactor empty round PET containers were attached to system was used in the study. The system consisted each bioreactor to serve as floats.The bioreactors of a round tapered plastic basket case (Figure 1) were subjected to 24/7 non-stop operation for 5 with a 20-mesh size net covering, engaged and months. fitted onto the basket’s periphery. An inner smaller plastic basket case was suspended and securely Collection of samples fastened at the center by cable ties and nylon Samples of raw water (pond water), water cords. Approximately 5.3 kilograms (21.95 dm3) of inside the tilapia cage and filtered water (bioreactor charcoal chips (1-5 cm dia.) were packed into the effluent) were collected from the Thai panga catfish main basket, filling all the spaces below 2 inches ponds monthly for 6 months. From the composite from the brim. A 600-L hr-1 capacity water pump water samples collected from each location or attached to a transparent plastic hose via sources, 1-liter sample was taken for water quality a t-connector which serves as the effluent output analysis and 30 mL for off-flavor (geosmin and lines was placed inside the smaller basket. MIB) analysis. Prior to bioreactor effluent collection, Finally, the surface of packed crushed charcoal the water outlet line was cleaned of algae and dirt was covered with a ring-shaped corrugated plastic and allowed to flow for 15 minutes. The collected

17 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

pump

Figure 1 An illustration of the charcoal bioreactor system used in the study samples in plastic bottles were immediately kept Samples was then heated to 65oC on a hotplate- in an ice box and transported to the laboratory stirrer and exposed to the SPME fiber for a 12-min within 24 hours. Off-flavor samples were stored in adsorption period while undergoing vigorous a freezer until ready for analysis. agitation. After 12 min, the fiber was withdrawn Phytoplankton was sampled every 30 days from the sample and desorbed on a splitless by filtration of 5-L composite water with a net of mode at 230oC for 5 min in the injection port of an 10 µm mesh. Samples were concentrated in HP 6890 N Network gas chromatograph equipped a 30-mL bottle and were preserved with Lugol’s with a 5973 mass selective detector (Agilent solution. Technologies, Palo Alto, CA). A Durabond HP-5 capillary column of 30-m length and 0.32-mm I.D. Analysis of off-flavor and 0.25-µm film thickness was used, with helium Water samples were collected from carrier gas operated at a rate of 2.5 ml/min. the experimental ponds and cages for geosmin The oven temperature was programmed at 6°C and MIB analyses. Off-flavor contamination in the for 1 min and increased to 220°C at a rate of samples will be analyzed by solid phase micro- 15°C min-1, maintaining the temperature at 220 extraction (SPME) and gas chromatography–mass degrees for 8 minutes. All analyses were performed spectrometry (GC-MS). Ten-milliliter sample of water as single measurements. Standard geosmin and was taken and transferred each to 20-mL straight- MIB from Sigma Company was used for calibration. sided headspace vials. Sodium chloride (1.9 g) and a PTFE-coated stirrer bar were added and the vial Water quality and nutrient analysis was sealed with an aluminum crimp cap-fitted Physico-chemical parameters were with a pre-pierced PTFE-faced silicone septum. measured in situ which include pH, temperature, The SPME fiber was extended into the headspace. dissolved oxygen and turbidity using a water

18 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28 quality multimeter (TOA Model WQC-22A, Japan). Olympus BH2 microscope with the aid of Laboratory analysis of total ammonia-nitrogen was a 1000_oil immersion objective. determined by phenate method, nitrate-nitrogen by cadmium reduction method, nitrite-nitrogen by Monitoring of fish growth diazotizing colorimetric method and phosphate- Samples of Nile tilapia and Thai hybrid phosphorus by stannous chloride method (Boyd panga catfish were randomly collected with and Tucker, 1992). specific fishing nets from each pond/cage every 30 days. The weight of the fish samples was Hydro-biological analysis determined using a weighing scale and was Chlorophyll-a was measured (Standard properly recorded. Method 10200H) on filtered (Whatman GF/C) water samples following extraction with 10 mL Data analysis hot methanol (60 o C in water bath), with Analysis of variance (ANOVA) using SPSS spectrophotometric detection of chlorophyll-a and (Version 16.0) for Windows was used to test for correction for phaeopigments and turbidity (APHA, difference between means of observed parameters 1995). Chlorophyll a concentrations in the extracts and each treatment. Post hoc analysis Duncan were calculated following Wintermans and Multiple Range Test (DMRT) at 95% confidence de Mots (1965) and Saijo (1975). Species and level was used for treatment comparison. T-test count of phytoplankton was determined using an was used to compare means between 2 groups.

Table 2 Growth performance of Nile tilapia and Thai hybrid panga

Treatment Parameter T1 T2 T3 STOCKING Total no. of tilapia 0 50 100 Ave. weight of tilapia (g) - 203.50+11.38 206.75+13.15 Total no. of Thai hybrid panga 300 300 300 Ave. weight of Thai hybrid panga (g) 230.15+8.53 212.25+47.75 208.75+23.03 HARVEST Ave. weight of tilapia (g) - 251.75+21.81 259.00+66.11

Ave. weight of Thai hybrid panga 672.25+121.98 623.75+86.8 593.75+9.22 SURVIVAL RATE (%) Nile tilapia - 84.50+9.15 82.00+2.45 a b a Thai hybrid panga 58.50+7.52 78.50+9.03 60.42+13.86 AVE. DAILY GROWTH RATE (g fish-1d -1) Nile tilapia - 0.30+0.10 0.31+0.47 Thai hybrid panga 2.66+1.00 2.48+0.59 2.31+0.08

19 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Results and Discussion panga was significantly higher in T2 than in T1 and T3. Treatment ponds had mean survival of Fish growth 78.50% (T2) and 60.42% (T3) while the controls The growth responses of both Thai panga had 58.50%. The major source of fish mortality and Nile tilapia during the 5-month experiment are in the ponds happened at the beginning of presented in Table 2. The number of fish in the experiment where some fish died due to the ponds decreased whilst fish biomass increased handling stress and were removed from the ponds in all treatments, indicating that individual fish without replacement. Besides mortality due to gained weight on average during the experiment. handling stress, survival of Thai panga remained Thai panga’s initial fish fresh weight significantly high, higher in T2 and T3 than in T1. Survival differed from the final weight in all three treatments rates of Nile tilapia were likewise high, which in (p<0.05). However, Nile tilapia in the cages (T2 fact much higher than that of Thai panga’s in the and T3) grew only slightly when fed exclusively three treatments. with phytoplankton. The fish significantly grew initially but lost weight during the 4th month and Effect on off-flavor occurrence recovered rapidly at the end of the experiment. Total off-flavor (GSM + MIB) concentrations This result is quite expected as in many other measured in water samples (pond water, cage enclosure experiments, fish either lost weight or water and bioreactor effluents) are shown in grew more slowly. Growth rates of Nile tilapia Figure 2. Water samples from the control and were comparable between T2 and T3 and were treated ponds contained significant levels of total not significantly related to density stocked (p>0.05). off-flavor concentration throughout the experimental Table 2 shows that doubling the density of Nile period (Figure 2). High concentrations in the control tilapia from 0.5m-2 to 1.0m-2 did not result in any treatment were recorded reaching as high as significant reduction in their mean daily growth 2.1 µgL-1 in August (Figure 3A). Total off-flavor rate (p>0.05), from 0.30g fish-1d-1 in T2 to 0.31g concentrations in T2 (low tilapia density) ranged fish-1d-1 in T3. This is in contrast with the study from 0.07–0.167 µgL-1 in pond water; 0.067–0.157 conducted by Attayde and Menezes (2008) which µgL-1 in cage water; and 0.05–0.127 µgL-1 in bioreactor demonstrated that growth rates of juvenile and effluent whilst total off-flavor concentrations in T3 adult Nile tilapia exclusively fed with phytoplankton (high tilapia density) ranged from 0.057–0.377 were inversely related to density stocked, indicating µgL-1 in pond water; 0.013–0.213 µgL-1 in cage strong density-dependent growth. Table 2 also water; and 0–0.183 µgL-1 in bioreactor effluent. shows the stocking and harvest data for the Total off-flavor levels (GSM+MIB) in pond waters study. There was no significant difference in mean of T2 and T3 were significantly lower (P<0.05) survival between tilapia (p>0.05) in T2 (84.50%) than T1 (control) (Figure 3B). However, off-flavor and T3 (82.00%) whilst mean survival of Thai levels were comparable between treatments

20 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

T2 and T3. Off-flavor was generally reduced ponds were dominated by green algae, Coelastrium in the pond waters of T2 and T3 through spp. and blue-green algae Microcystis spp. during the interventions, despite of some inconsistencies the experimental period. Treatment ponds at low with the trend: pond water > cage water > bioreactor tilapia density (T2) were dominated by Microcystis effluent (Figure 3C-3D). spp. and Dictyosphaerium spp. whilst those ponds with high tilapia density (T3) were dominated by Reduction of cyanobacteria in ponds Coelastrium spp. and Eudorina spp. Examinations Algae, including 77 genera in six of algal species composition revealed considerable phyla were identified in the twelve fishponds. reduction of cyanobacteria (Figure 4A-4B) in both The 77 genera belonged to Chlorophyta (40), cage water and bioreactor effluents relative to Cyanophyta (12), Baccillariophyta (13), Charophyta (7) pond water in treatment ponds T2 and T3. Euglenophyta (4) and Cryptophyta (1). The control

A B

C D

Figure 3 (A-D) Total off-flavor [geosmin (GSM) + 2-methylisoborneol (MIB)] concentration in pond water, cage water and bioreactor effluent in control (T1) and in low (T2) and high (T3) tilapia density integrated-culture treatments

21 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Cyanobacterial cell density in both treatments can reduce chlorophyll a and algae or cyanobacteria decreased in the following order: pond water > (and off-flavor) in the pond water. Nevertheless, in cage water > bioreactor effluent which suggests both T2 (low tilapia density) and T3 (high tilapia effectiveness of the intervention. Twelve genera of density), no significant differences (P>0.05) in cyanobacteria were identified in T3 and nine chlorophyll a values were found among pond, genera in T2. Among these, Anabaena spp. and cage and bioreactor effluent water. This was Oscillatoria spp. and Synechococcus spp. (Smith probably due to the size structure of the phytoplankton et al., 2008) are known producers of off-flavor community in the experimental ponds. According compounds (geosmin and 2-methylisoborneol) in to Drenner et al. (1987), particle ingestion by filter- water as well as Microcystis spp. which some feeding tilapia is a function of particle size and species were implicated to off-flavor problems greatly decreases when a particle diameter is (Watson, 2003; Zhong et al., 2011). Species <20–30 µm. In short, tilapia filter feeders directly composition of cyanobacteria was higher in the suppress populations of less mobile zooplankton high tilapia density treatment compared to the low and large phytoplankton (cyanobacteria), and density treatment. indirectly enhance small algae (nanoplankton). The average amounts of chlorophyll-a in The comparable concentration of chlorophyll a in the treatment ponds were generally higher over the bioreactor’s effluent water with the pond/cage time compared to the control ponds except in water could be due to these small algae that were August (Figure 5). All of the ponds behaved the continuously sucked in through the bioreactor same way displaying similar trends regardless of (passing through the 20-mesh size net covering) treatment. While there was a slight decline in and eventually released in the effluent. The inability average chlorophyll-a values for treated ponds of the bioreactor to reduce chlorophyll a (possibly from March to June, the control pond experienced due to nanoplankton) and increase water transparency a slight rise in chlorophyll a from 62.94 µgL-1 to (low turbidity) could be due to its very low hydraulic 107.57 µgL-1. There was a significant increase in retention time (HRT) of 0.2 hour (9.8 min). HRT is the average chlorophyll-a levels from June defined as the bioreactor capacity (21.95 L) (102.89–133.13 µgL-1) to August (445.00–610.67 divided by flow rate (2.25 L min-1). A very recent μgL-1) and a subsequently decreased towards study showed that the optimum HRT (bio-fence September in all treatments. Nile tilapia, both as type charcoal bioreactor) to reduce chlorophyll a of a visual and filter feeder, can reduce excessive an aquaculture pond by 80% was at 6 hours phytoplankton biomass and recycle nutrients (Bi, 2013) which translates that the flow rate of effectively (Stickney et al., 1979). It was hypothesized the charcoal bioreactor in this study should only that the combined intervention of biomanipulation be around 0.61 L min-1. with Nile tilapia and the use of charcoal bioreactor

22 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

B A

Figure 4 (A-B) Cyanobacterial composition in low (T2) and high (T3) tilapia density integrated-culture

Figure 5 Variations in chlorophyll a in different treatments over the 5-month experimental period

However, the bioreactor shows its effectiveness were generally similar inside and outside the tilapia in reducing cyanobacterial biomass. Moreover, the cage and in the effluent in T2 and T3 (Figure 6A). effects of Nile tilapia on phytoplankton community No significant difference in pH among the control should depend on the fish biomass (Starling et al., and treated ponds was observed (p>0.05) (Figure 1998) but no effects of a 2-fold increase in Nile 6B). The same result was observed among pond, tilapia biomass were found in this study. cage and bioreactor effluent water within the two treated ponds, T2 and T3. In the tilapia integrated Water quality parameters ponds the general pattern was for an initial Surface temperature of pond water, cage decrease followed by a gradual increase and water and bioreactor effluent ranged from 31.27– eventually decreases gain until the end of each 34.93◦C at the start of sampling in March, and experimental period. Water transparency (Figure from 29.63–30.23◦C at the end of experiment in 6C) was not affected significantly by the presence September. The mean values of temperatures of tilapia and bioreactor use, showing comparable 23 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28 turbidity (p>0.05) in all treatments. The apparent Generally, the nutrient concentration increase in water turbidity (including that of the control fluctuated over time during the study period. ponds) towards the end of the experimental period Treatments had different effects on nutrient levels. could be due to the corresponding increase in Total ammonia-nitrogen (TAN) in the control chlorophyll-a concentration. Also, the inflows from ponds followed a similar trend as in the treatment localized run-offs due to rains could have increased ponds (Figure 6E). However, a gradual decline the turbidity of the pond water during the rainy was observed in the tilapia integrated treatments months of August and September. over the first three months, followed by a significant Although dissolved oxygen concentration increase to the end of the experiment reaching did not change in the cages in relation to the pond a maximum concentration above 3 mgL-1 in the low in the integrated-culture treatments, it was consistently (T2) and high (T3) density experiments. The mean different in the bioreactor effluents (Figure 6D). In increases in ammonium in T2 and T3 were 92% the latter DO was consistently lower than in cage and 34%, respectively. Despite the high TAN values, and pond water, though the levels of DO recorded no fish mortality was recorded which suggests (lowest was at 4 mgL-1 in T2 bioreactor) were still that most of the TAN were in the ionized form -1 o way above the threshold value of 1mgL at 26 C (NH4-N). The concentration of the unionized form, for Nile tilapia (Duy et al., 2008). DO gradually NH3, which is toxic to fish (both freshwater and decreases except for an intermittent increase in marine) at >0.03 mg/L, depends on the pH and the latter months recorded in pond and cage water temperature of water. There was an overall increase in all the treatments. Measured DO in the control in average nitrate levels in all treatments except in ponds were higher than in T2 and T3 particularly the control ponds (Figure 6F). This rise in nitrate those measured in May and August. The reduced concentration might be due to the amount of fish oxygen concentrations may have been due in part excrement that is added in the form of ammonia to the respiration of the increased fish biomass which is later converted to nitrate. It was considerably (both tilapia and panga) in or to the increased enhanced by the presence of tilapia particularly in turbidity (Figure 6C), which could reduce high density treatment (T3) reaching values as photosynthesis. The charcoal bioreactor used high as 1.5 mgL-1. Nitrate levels for all of the treated a 600-L hr-1 water pump and maintained an average ponds were higher than in the control ponds flow rate of 2.25 Lmin-1 to help circulate water in throughout the study except initially for the month the cage and prevent hypoxia inside the cage of March where control had a higher average especially at dawn when low DO levels are value. Nitrite was also affected significantly by usually observed. However, high flow rates decrease the presence of Nile tilapia in T2 and T3, showing the HRT and can affect the bioreactor performance an increasing trend over time until August but in removing smaller algae in the water as observed decreases at the end of the experiment (Figure in the present study. 6G). Conversely, nitrite levels in the control ponds

24 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28 were consistently lower than in T2 and T3 except enhanced by the presence of Nile tilapia in T2 and in September where average value was higher. T3 (Figure 6H). Based on the results, biomanipulation

Concentrations of orthophosphate in T2 and T3 using Nile tilapia affected nutrients. TAN, NO3-N, increased in the middle of the experimental period NO2-N and PO4-P concentrations were generally and a slight reduction in the final months was higher in T2 and T3 relative to non-integrated culture observed. Similarly, orthophosphate was slightly ponds.

A B

C D

E F

G H

Figure 3 (A-H) Water quality of pond water, cage water and bioreactor effluent 25 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Though Nile tilapia reduced algal population by Reference grazing as shown in this study, it could also lead to ichthyo-eutrophic condition as the results APHA. 1995. Standard methods. 19th Edition. suggested. Datta-Saha and Jana (1998) observed Washington, DC: American Public Health that about 6 to 180% nutrient enrichment was Association. attributable to the defecation and excretion of the Attayde, J.L. and R.M. Menezes. 2008. Effects test fishes. Further, results also suggest that the of fish biomass and planktivore type on charcoal bioreactor used in this study had no plankton communities. J. Plank. Res. effect on water quality. 30: 885-892. Bi, X. 2013. Study on bioreactor system using Conclusion charcoal for purification of toxic cyanobacteria in eutrophic water Aquaculture systems either mono-culture bodies. MS Thesis. Nagasaki University. or integrated, employing the green water culture Bowen, S.H. 1983. Feeding, digestion and system are inevitably subject to off-flavor problems. growth, qualitative considerations. In: The combined intervention of biomanipulation using Pullin, R.S.V. (eds) The Biology and Nile tilapia, and the use of charcoal bioreactor Culture of Tilapias. Lowen-McConnell. (basket-type) in controlling cyanobacteria and off- pp. 141-156. flavor in water appears to be quite effective, Boyd, C.E. and C.S. Tucker. 1992. Water though inconclusive in some aspects and needs Quality Pond and Soil Analyses further study. The effectiveness and contribution for Aquaculture, Alabama Agriculture of the charcoal bioreactor in reducing off-flavor in Experiment Station. Auburn University, the pond is still not clear due to the observed Auburn (AL). 62 p. comparable differences in off-flavor levels in water Carpenter, S.R. and J.F. Kitchell. 1992. Trophic within the experimental treatments. However, the cascade and biomanipulation: interface of study demonstrates that Nile tilapia can be cultured research and management–a reply to the effectively with Thai hybrid panga, establishing comment by DeMelo et al. Lim. Ocean. the economic benefits of the integrated culture, as 37: 208-213. well as the effective control of cyanobacteria and Chapman, G. and C.H. Fernando. 1994. off-flavors. The diets and related aspects of feeding of Nile tilapia (Oreochromis niloticus L.) Acknowledgements and common carp (Cyprinus carpio L.) in lowland rice fields in northeast The authors thank the staff of Thai Panga Thailand. Aquaculture. 123: 281-307. Farm (TPF) in Kalasin Province especially to Ms. Sarinya Mulnee and the National Research Council of Thailand for a research grant. 2826 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Chen, S.L., S.F. Liu, C.L. Hu and L. Tian. 1990. Itayama, T., N. Iwami, M. Koike, T. Kuwabara., On the digestion and utilization of N. Whangch and N. Inamori. 2008. Microcystis by fingerlings of silver carp Measuring the effectiveness of a pilot and bighead. Acta Hydrobiologica scale bioreactor for removing Microcystis Sinica. 14: 49-59. in an outdoor pondsystem. Envi. Sci. Datta, S. and B.B. Jana. 1998. Control of bloom and Technol. 42: 8498-8503. in a tropical lake: grazing efficiency of Jančula, D., Drábková, M., Adámek, Z. and some herbivorous fishes. J. Fish Biol. Maršálek, B. 2008. Changes in the 53: 12-24. photosynthetic activity of Microcystis Drenner, R.W., G.L. Hambright, G.L. Vinyard. colonies after gut passage through Nile 1987. Experimental study of size- tilapia (Oreochromis niloticus) and silver selective phytoplankton grazing by a filter- carp (Hypophthalmichthys molitrix). feeding cichlid and the cichlid’s effects on Aqua. Res. 39: 311-314. plankton community structure. Jin, C., S. Dong, B. Gu and S.H. Bowen. 2006. Limnol. Oceanogr. 32: 1138-1144. Feeding and control of blue-green algal Duy T., J. Scharama, A.V. Dam, and A.J. Verreth blooms by tilapia (Oreochromis niloticus). 2008. Effects of oxygen concentration Hydrobiologia. 568: 111-120. and body weight on maximum feed Kolmakov, V.I. and M.I. Gladyshev. 2003. intake, growth and hematological of Nile Growth and potential photosynthesis of tilapia, Oreochromis niloticus. cyanobacteria are stimulated by viable gut Aquaculture. 275: 152-162. passage in crucian carp. Aq. Ecol. Friedland, K.D., D.W. Ahrenholz and L.W. Haas. 37: 237-242. 2005. Viable gut passage of Komarkova, J. 1998. Fish stock as a variable cyanobacteria through the filter-feeding modifying trophic pattern of fish Atlantic menhaden Brevoortia phytoplankton. Hydrobiologia. tyrannus. J. Plank. Res. 27: 715-718. 369/370: 139-152. Gavel, A., B. Marsalek and Z. Adamek. 2004. Lu, J., J.T. Takeuchi and H. Satoh. 2006. Viability of Microcystins colonies is not Ingestion, assimilation and utilization of damaged by silver carp raw Spirulina by larval tilapia Oreochromis (Hypophthalmichthys molitrix) digestion. niloticus during larval development. Algological Studies. 113: 189-194. Aquaculture. 254(1-4): 686-692. Getachew, T. 1989. Stomach pH, feeding Miura, T. Miura and J. Wang. 1985. Chlorophyll a rhythm and ingestion rate in Oreochromis- found in feces of phytoplanktivorous Niloticus L. (Pisces, Cichlidae) in Lake cyprinids and photosynthetic acitivity. Awasa, Ethiopia. Hydrobiologia. Verh. Internat. Verein. Limnol. 174: 43-48. 22: 2636-2642.

27 Journal of Agr. Research & Extension 30(3) (Suppl.): 14-28

Moriarty, D.J.W. 1973. The physiology of Watson, S.B. 2003. Cyanobacterial and digestion of blue-green algae in the eukaryotic algal odour compounds-signals cichlid fish, Tilapia nilotica. J. Zoo. or by-products? A review of their bio- 171: 25-39 logical activity. Phycologia. 42: 332-350. Saijo, Y. 1975. A method for determination of Wintermans, J.F.G.M. and A. de Mots. 1965. chlorophyll. Jpn J. Limnol. 36: 103-109. Spectrophotometric characteristics of Shapiro, J 1990. Biomanipulation: the next phase chlorophyll a and b and their pheophytins –making it stable. Hydrobiologia. in ethanol. Biochim. Biophys. Acta. 200/201: 13-27. 109: 448-453. Smith, J.L. G.L. Boyer and P.V. Zimba. 2008. Xie, P. 2003. Silver Carp, Bighead and Control A review of cyanobacterial odorous and of Algal Blooms. Beijing (in Chinese). bioactive metabolites: impacts and Science Press. 19-69. management alternatives in aquaculture. Xie, P. and J.K. Liu. 2001. Practical success of Aquaculture. 280: 5- 20. biomanipulation using filter-feeding fish to Starling, F., M. Beveridge and X. Lazzaro. 1998. control cyanobacterial blooms: a synthetic Silver carp biomass effects on the of decades of research and application in plankton community in paranoa reservoir a subtropical lake. The Scientific World. (Brazil) and an assessment of its potential 1: 337-356. for improving water quality in lacustrine Zhong, F.Y. Gao, T. Yua, Y. Zhang, D. Xua, E. Xiao, environments. Int. Rev. Hydrobiol. F. He, Q. Zhoua and Z. Wua. 2011. The 83: 499-507. management of undesirable Starling, F.L.R.M. 1993. Control of cyanobacteria blooms in channel catfish eutrophication by silver carp ponds using a constructed wetland: (Hypophthalmichthys molitrix) in the Contribution to the control of off-flavor tropical Paranoa Reservoir (Brasilia, occurrences. Water Res. 45: 6479-6488 Brazil): a mesocosm experiment. Hydrobiologia. 257: 143-152. Stickney, R.R. J.H. Hesby, R.B. Mcgeachin and W.A. Isbell. 1979. Growth of Tilapia niloticus in ponds with differing histories of organic fertilization. Aquaculture. 17: 189-194.

28 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Medicinally Potential Plant of Anisomeles malabarica (L.) R. Br.

Rameshprabu Ramaraj1,2* and Yuwalee Unpaprom3 1Division of Entomology, Department of Zoology, School of Life Sciences, Bharathiar University, Coimbatore-641 046, Tamil Nadu, . 2School of Renewable Energy, Maejo University, Chiang Mai, Thailand 50290 3Department of Biology, Faculty of Science, Maejo University, Chiang Mai, Thailand 50290 *Corresponding author: [email protected]

Abstract

Anisomeles malabarica (L.) R. Br. is perhaps the most useful traditional medicinal plant. It is a highly aromatic plant belonging to the family (Labiatae). Anisomeles malabarica is a species of herbaceous plant native to tropical and subtropical regions. Mosquitoes act as a vector for most of the life threatening diseases like malaria, yellow fever, dengue fever, filariasis, encephalitis. Part of the present study was aimed to evaluate the larvicidal and pupicidal activities of crude methanol extract of Anisomeles malabarica. The extract was assayed for their toxicity against the important vector mosquitoes Anopheles stephensi. The plant extract showed larvicidal effects after 24 h of exposure; however, the highest larval and pupal mortality was found in the methanol extract of Anisomeles malabarica against the first to fourth instars larvae and pupae. This result suggested that the plant extract have the potential to be used as an ideal eco-friendly approach for the control of mosquito vector. This paper gives a bird’s eye view mainly on the biological activities, pharmacological actions, and plausible medicinal applications of Anisomeles malabarica. Different aspects of Anisomeles malabarica medicinal values are briefly demonstrated, such as potential anti-allergic, anti-anaphylactic, anti-bacterial, anticancer, anti-carcinogenic, anti-inflammatory, antiepileptic potential, antifertility, anti-pyretic activity and antispasmodic.

Keywords: Anisomeles malabarica, medicinal , medicinal significance

Introduction Medicinal plants are the principal health care resources for the majority of people throughout Medicinal plants have long history as the world. The plant kingdom has provided a vast important components in traditional medicine, and source of medicinal plants first used in their crude food of humans since the times of ancient Indians, forms as herbal teas, syrups, infusions, ointments, Egyptians, Chinese and many other countries liniments and powders (Samuelsson, 2004). (McCurdy and Scully, 2005; Sneader, 2005).

29 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Asia is the largest continent, with 60% of nutlets, bearing ellipsoid and compressed seeds the world’s population. The region consists of (Joshi, 2000; Singh et al., 2003). The common the continent of Asia including the islands in vernacular names of the plant are Peimiratti, the Indian and the Pacific Oceans. There are Malabar catmint, Bhutan kusham and Peyameratti. abundant medicinal and aromatic plant species and The taxonomic classification of A. malabarica is traditional medicine which been used since ancient as follows: times (Gurib-Fakim, 2006). However, the biodiversity of medicinal and aromatic plants is yet to be studied Division Spermatophyta thoroughly in many countries. A very important Sub-division Angiospermae medicinal plant family is Labiatae, also known as Class Dicotyledone the mint family. The Labiatae family (Lamiaceae) Sub-class Gamopetalae is one of the largest and most distinctive families Series Bicarpellate of flowering plants, with about 220 genera and Order almost 4,000 species worldwide. Plants in this Family Labiatae or Lamiaceae family are herbs or often with an aromatic Genus Anisomeles smell. They are common in the Asia, Maltese Species Malabarica R.Br. Islands and other Mediterranean countries. Labiatae Botanical name Anisomeles malabarica family is well represented; many members of this (L.) R. Br. family are used in traditional and folk medicine. They are also used as culinary and ornamental plants such as mints, thyme, tulsi, spearmint and coleus (Naghibi et al., 2005). Medicinal constituents include the strong aromatic essential oil, tannins, saponins and organic acids. The Labiatae plants have sedative, diuretic, tonic, antispasmodic and antiseptic properties. Anisomeles malabarica (L.) R.Br. (A. malabarica) has been provided also many important uses. A. malabarica is an aromatic, densely pubescent, perennial herb, 1.2-2.0 meter in height. Leaves are simple, opposite, very thick, aromatic, oblong-lanceolate, acute, pale above, white below, crenate-serrate, and woolly; flowers purple, in dense whorls of more or less interrupted spikes; fruits Figure 1 Anisomeles malabarica

30 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Earlier phytochemical studies of A. malabarica medicinal values of A. malabarica plant details have shown the presence of anisomelic acid, were shortly demonstrated. anisomelolide, 2-acetoxymalabaric acid, anisomelyl acetate anisomelin, betulinic acid, ß-sitosterol, Citral, Materials and methods gerainic acid, malabaric acid, ovatodiolide, and triterpenebetulinic acid (Guha Bakshi et al., 1999; The methodology is illustrated in (Figure Choudhary et al., 2001). 2). A. malabarica was appeared to have the best Since A. malabarica is a medicinal plant, source of medicinal value in different aspects. has been used as a folkloric medicine to treat amentia, anorexia, fevers, swellings, rheumatism. Plant collection and identification The plant A. malabarica is used traditionally in the The whole plant of A. malabarica, were treatment of intermittent fever, colic dyspepsia collected from around the Bharathiar University and curing wounds (Jeyachandran et al., 2007). campus, Coimbatore, Tamil Nadu, India. It was In some cases, the crude extract of identified at Department of Botany, Bharathiar medicinal plants may be used as medicaments. University, Coimbatore, Tamil Nadu, India and Medicinal plants can provide biologically active the voucher specimens were deposited of molecules and lead structures for the development the Entomology Division, Zoology Department, of modified derivatives with enhanced activity. Bharathiar University. Some naturally occurring botanical compounds contain a broad range of chemical active ingredients Plant extraction which can intervene in all biological processes of The leaves and were the mosquito, thus interrupt its life cycle and washed with tap water and shade dried at room dispersal and reduce harms to humans and temperature. The dried plant material was powdered animals. Many medicinal plants were tested for mechanically using commercial electrical stainless their pesticide and repellent potential, as crude steel blender. The powdered plant material was material, essential oils or individual active ingredients successively extracted with methanol (80ºC) by (Ramaraj, 2005). Consequently, this paper highlight hot continuous percolation method in Soxhlet the above, provides an overview of the classes of apparatus for 24 hrs (Vogel, 1978).The extracts molecules present in plants and gives some filtered through a Buchner funnel with Whatman examples of the types of molecules and secondary number 1 filter paper. The extract was further metabolites that have led to the development of concentrated by recovering excess solvents to these pharmacologically active extracts. Therefore, thick oily natured crude in a rotary evaporator at this study also presents some data on the reduced pressure and subjected to freeze drying significance of A. malabarica against malarial in a lyophilizer till dry powder was obtained. vector (Anopheles stephensi). In addition, the

31 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Mosquito larval toxicity test needle and moribund larvae were counted as

A laboratory colony of Anopheles stephensi dead. The LC50 and LC90 value was calculated (An. stephensi) larvae was used for the larvicidal after 24 h by probit analysis (Finney, 1971). activity (Ramaraj, 2005). Twenty five of first, second, third and fourth instar larvae were kept in 500 ml Mosquito pupal toxicity test glass beaker containing 249 ml of dechlorinated A laboratory colony of Anopheles stephensi water and 1ml of different concentration of plant pupae were used for pupicidal activity. Twenty of extracts. Larval food was given for the test larvae. freshly emerged pupae were kept in 500 ml glass At each tested concentration 2 to 5 trails were beaker containing 249 ml of dechlorinated water made and each trial consisted of three replicates. and 1 ml of desired plant extract at various The control was set up by mixing 1 ml of acetone concentrations was added. Five replicates were with 249 ml of dechlorinated water. The larvae set up for each concentration and control was exposed to dechlorinated water without acetone setup by mixing 1ml of acetone with 249 ml of served as control. Larval mortality was assessed dechlorinated water. The control mortality was after 24 h of exposure by probing the larvae with corrected by Abbott’s formula (Abbott, 1925).

Anisomeles malabarica Medicinal value

Plant collection & Identification Anti-allergic

, Anti-anaphylactic , Anti-bacterial Anti-pyretic activity Antispasmodic Preparation of plant extract Anticancer Anti-carcinogenic Anti-inflammatory Mosquito culture & control Antiepileptic potential Antifertility Anti-pyretic activity Malaria, dengue etc. Antispasmodic

Figure 2 Methodology and medicinal value of Anisomeles malabarica

32 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

The average larval and pupal mortality data were high morbidity and mortality for many patients subjected to probit analysis for calculating, LC50, (Viroj, 2011). Since ancient times, several plants LC90 and other statistics at upper confidence limit and plant products have been used locally to (UCL) and lower confidence limit (LCL) and chi- repel or kill or control mosquitoes which case square values calculated using the SPSS 9.0 fever and malaria (Curtis et al., 1991; Berger and version (software package). The values were Curtis, 1995). expressed as mean ± standard deviation of five The highest parasite mortality was found replicates. Results with p<0.05 were considered in leaf methanol extracts of A. malabarica (LC50 = th to be statistically significant. In addition, all data 466.15 ppm and LC90 = 1837.96 ppm) against the 4 were subjected to analysis of variance (ANOVA), instar larvae of Anopheles subpictus (Zahir et al., completely Randomised Design (CRD) and the 2009). means were separated using Duncan’s multiple In this study, Larval and pupal toxicity range test (DMRT) (Alder and Rossler, 1977; effect of A. malabarica leaf extract (AMLE) on An. Ramaraj, 2005). stephensi is shown in Figure 3. Similar trend has been noted for all the instars of An. stephensi at Results and discussion different concentrations of AMLE. The LC50 and LC90 values were as follows: LC50 value of I instar Dengue fever and malaria was 50.24%, II instar was 54.70%, III instar was Dengue fever and malaria are the two 59.03% and IV instar was 64.33%, respectively. common mosquito infections in tropical and sub- LC90 value of I instar was 106.99%, II instar was tropical regions that are very important and cause 113.33%, III instar was 118.04% and IV instar

100

80 I 60 II III 40 IV % of % mortality Pupa 20

0 20 40 60 80 100 Concentration (ppm)

Figure 3 Larval and pupal toxicity of ethanolic extract of Anisomeles malabarica leaf on Anopheles stephensi 33 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39 was 119.94%, respectively. At 20 ppm of AMLE Complete reduction of larva was noticed treatment the pupal mortality was 23% increased after combination of AMLE and AMIFE. The toxicity to 82% at 100 ppm with AMLE treatment. The LC50 of methanol extracts of A. Malabarica on survival value after the treatment of AMLE was 55.07% of An. Stephensi reveals that the mortality is and LC90 value was 117.89%, respectively. dependent on the dose of the plant extract. Both The larval and pupal toxicity effect of extracts had detrimental effect on larval and pupal Anisomeles malabarica inflorescence extract growth and development. (AMIFE) on An. Stephensi is shown in Figure 4. At 20 ppm I instar the larval mortality was 15%, Anti-allergic whereas at 100 ppm the mortality increased The increase of allergic diseases has to 84%. The LC50 and LC90 values were as accompanied the global population growth and follows: LC50 value of I instar was 64.76%, II the major challenge is to reduce morbidity. Many instar was 67.92%, III instar was 71.46% and IV natural products have been identified as potential instar was 74.55%, respectively. LC90 value of I anti-allergic agents. In addition, plant formulations instar was 118.22%, II instar was 121.60%, III have demonstrated, in general, to be safe in instar was 123.65% and IV instar was 119.94%, clinical trials and demonstrate additional effects respectively. At 20 ppm concentration the pupal along with Western medicines such as synergism mortality was increased to 78%. The LC50 value and modulation of the immune system (Cota after the treatment of AMIFE was 61.67% and et al., 2013). Essential oil of A. malabarica was LC90 value was 122.89%, respectively (Figure 4). found to have anti-allergic properties (Jeyachandran et al., 2007).

100

80 I

60 II III

% of mortality mortality of % 40 IV 20 Pupa

0 20 40 60 80 100 Concentration (ppm)

Figure 4 Larval and pupal toxicity effect of ethonalic extract of Anisomeles malabarica inflorescence, Anopheles stephensi 34 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Anti-anaphylactic Anticancer Anaphylaxis is a serious allergic reaction Cancer is one of the most severe health that is rapid in onset and may cause death. problems in both developing and developed Anaphylaxis is a severe, potentially fatal, systemic countries, worldwide. Medicinal plants with their allergic reaction that occurs suddenly after contact isolated lead molecules are also used as an with an allergy-causing substance. Worldwide 0.05- alternative medicine for treating neoplastic cells 2% of people are estimated to have anaphylaxis which are the anomalous proliferation of cells in at some point in their life and the occurrence rate the body which cause cancer. Diverse efficient appears to be increasing (Tintinalli, 2010). Currently, compounds derived from natural products have focus on plant research has improved all over been isolated as anticancer agents. These chemical the world and a large body of evidence is accumulated compounds are formulated with a view to create to show immense potential of medicinal plants effective drugs against cancer. Chopra et al. (1956) used in various traditional systems. A. malabarica stated a number of pharmacological effects has essential chemical compounds serve as an of A. malabarica. Jeyachandran et al. (2007) anti-anaphylactic agent (Kavitha et al., 2012). reemphasized the usefulness of A. malabarica in traditional medicine as an anticancer agent used Anti-bacterial in traditional medicine. Bacterial pathogens have evolved numerous defense mechanisms against antimicrobial agents; Anti-carcinogenic hence resistance to old and newly produced drugs is Carcinogens are involved in carcinogenesis, on the rise. The phenomenon of antibiotic resistance mutagenesis and genotoxicity. One of the best exhibited by the pathogenic microorganisms has led ways to minimize the detrimental effects of to the need for screening of several medicinal carcinogens is by the use of medicinal plants i.e., plants for their potential antimicrobial activity. natural anticarcinogens. These include flavonoids, Chemical compounds from the leaves of A. phenolics, coumarins, carotenoids, anthraquinones, malabarica have a wide spectrum of antibacterial tannins, saponins etc. These natural products play action against Gram-negative and Gram-positive important role, not simply in the primary prevention pathogenic bacteria such as Escherichia coli, of cancer, but also in the prevention of cancer Klebsiella pneumoniae, Staphylococcus aureus, recurrence, which is of greatest importance in Vibrio cholera, pseudomonas aeruginosa, and determining survival (Sanjib, 2012). A. malabarica Proteus mirabilis (Kavitha et al., 2012; Remya was found to have tremendous anti-carcinogenic et al., 2012). agents (Kavitha et al., 2012).

35 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Anti-inflammatory Anti-fertility Anti-inflammatory refers to the property of Anti-fertility is the prevention of conception a substance or treatment that reduces inflammation. or impregnation. Numerous herbs have been used Anti-inflammatory drugs make up about half historically to reduce fertility and modern scientific of analgesics, remedying pain by reducing research has confirmed anti-fertility effects in at inflammation as opposed to opioids, which affect least some of the herbs tested. Setty et al. (1976) the central nervous system. confirmed that A. malabarica has antifertility A. malabarica has significant anti- potential. inflammatory activity which may be due to presence of chemical profile such as flavones, tri- Anti-pyretic activity Terpenoids, flavonones and phenols (Lavanya Antipyretics are drugs or herbs that reduce et al., 2010). fever. Antipyretics cause the hypothalamus to override an interleukin-induced increase Antiepileptic potential in temperature. The body then works to lower The antiepileptic (also commonly known the temperature, resulting in a reduction in fever. as anticonvulsants drugs) is a diverse group of The various plants (such as Adansonia digitata, pharmaceuticals used in the treatment of epileptic Berberis species, Capparis zeylanica, Cleome seizures. Anticonvulsants are also increasingly viscose, Vernonia cinerea etc.) used as an antipyretic used in the treatment of bipolar disorder. agents (Shah and Seth, 2010; Srinivasan et al., Medicinal plants used for the therapy of epilepsy 2010). in traditional medicine have been shown to possess promising anticonvulsant activities in Antispasmodic animal models of anticonvulsant screening can be An antispasmodic is a type of medication an invaluable source for search of new antiepileptic that is primarily used to treat convulsions or compounds. A. malabarica has numerous therapeutic uncontrollable muscle movements, particularly in utilities in folk medicine. The ethyl acetate extract the intestines and stomach. The drug is also of A. malabarica leaves. It was founded that the commonly referred to as phenobarbital, which extract of A. malabarica leaves has antiepileptic belongs to the barbiturate drug group. A. malabarica capability (Choudhary et al., 2011). has crucial chemical compounds which are produce antispasmodic agent (Jeyachandran et al., 2007).

36 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Conclusion Alder, H.L. and E.B. Rossler. 1977. Introduction to probability and Anisomeles malabarica, the versatile statistics 6th edition. San Francisco: medicinal plant is the unique source of various W.H. Freeman Company. 246 p. types of compounds having diverse chemical Berger, A. and C.F. Curtis. 1995. Natural plant structures. Mosquito-borne diseases have economic products as pesticides. In Swedish impact including loss in commercial and labor University of Agricultural Sciences. outputs; particularly in countries with tropical and Alnarp: Swedish University of Agriculture subtropical climates. The current investigation Science. 45 p. revealed that the crude extract of A. malabarica Chopra, R.N., S.L. Nayar and I.C. Chopra. 1956. possesses remarkable larvicidal and pupicidal Glossary of Indian Medicinal Plants. activities against the mosquito vectors. The result Council of Scientific and Industrial shows that good larvicidal properties against Research. New Delhi. vector control programs. Higher larval and pupal Choudhary, N., K.R. Bijjem and A.N. Kalia. 2011. mortality was shown after the treatment A. malabarica Antiepileptic potential of flavonoids inflorescence and leaves extracts. These results fraction from the leaves of Anisomeles suggest that the inflorescence and leaves extracts malabarica. J. Ethnopharmacol. have the potential to be used as an ideal eco- 135(2): 238-242. friendly approach for the control of mosquitoes. Cota, B.B., C. M. Bertollo and D.M. de Oliveira. A. malabarica has numerous therapeutic utilities in 2013. Anti-allergic potential of herbs and folk medicine. Ethno botanically, the leaves of this herbal natural products-activities and plant are used against convulsions, for dyspepsia in patents. Recent Pat Endocr Metab intermittent fevers, colic, boils, tetanus, inflammation, Immune Drug Discov. 7(1): 26-56. cough, cold, stomachache, itches and in uterine Curtis, C.F., J.D. Lines, L. Baolin and A. Renz. affections. This study also helps to differentiate 1991. Natural and synthetic repellents. the valuable A. malabarica for phyto-pharmaceuticals pp. 75-92. In Control of disease vectors and medicinal values. in the Community. London: Wolf Publisher Limited. Reference Finney, D.J. 1971. Probit analysis. London: Cambridge University Press. 272 p. Abbot, W.S. 1925. A method of computing Gurib-Fakim, A. 2006. Medicinal plants: the effectiveness of an insecticide, Tradition of yesterday and drugs of J. Econ. Entomol. 18: 265-267. tomorrow. Review article. Mol. Aspects Med. 27(1): 1-93.

37 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

GuhaBakshi, D.N., P. Sensarma and D.C. Pal. Ramaraj Rameshprabu. 2005. Studies on the 1999. A Lexicon of Medicinal Plants in larvicidal, pupicidal and repellent India. Calcutta: NayaPrakash; properties of Anisomeles malabarica (L) 206. 152 p. for the control of malaria vector Joshi, S.G. 2000. Medicinal Plants. NewDelhi: Anopheles stephensi Listen. M.Phil. Mohan Primlonefor Oxford and IBH Thesis. Bharathiar University. Nadu. Publishing Co Pvt. Ltd. 222 p. Remya, M., Pankaj Jha and Someshwar Nath. Jeyachandran, R., A. Mahesh and L. Cindrella, 2012. Bioactivity studies on Anisomeles 2007. DEN-Induced cancer and its malabarica (AM) R.Br. J. Biotech. and alleviation by Anisomeles malabarica (L.) Biotherapeutics. 2(9): 1-8. R.Br. ethanolic leaf extract in male albino Samuelsson, G. 2004. Drugs of natural origin: mice. Int. J. Cancer Res. 3: 174-179. a textbook of pharmacognosy 5th Kavitha, T., R. Nelson, R. Thenmozhi and Stockholm: Swedish Pharmaceutical Press. E. Priya. 2012. Antimicrobial activity and Sanjib Bhattacharya. 2012. Anticarcinogenic phytochemical analysis of Anisomeles property of medicinal plants: involvement malabarica (L) R. BR. J. Microbiol. of antioxidant role. pp. 83-96. In Biotech. Res. 2(1): 1-5. Capasso A. (ed.) Medicinal plants as Lavanya, R., S. Uma Maheshwari, G. Harish, antioxidant agents: understanding their J. Bharath Raj, S. Kamali, D. Hemamalani, mechanism of action and therapeutic J. BharathVarma, C. Umamaheswara efficacy. Research Signpost, Trivandrum. Reddy. 2011. Investigation of In-vitro India. anti-Inflammatory, anti-platelet and Setty, B.S., V.P. Kamboj, H.S. Garg and N.M. anti-arthritic activities in the leaves of Khanna. 1976. Spermicidal potential of Anisomeles malabarica Linn. RJPBCS. saponins isolated from Indian 1(4): 745-752. medicinal plants. Contraception. McCurdy, C.R. and S.S. Scully. 2005. Analgesic 14: 571-578. substances derived from natural products Shah, B.N. and A.K. Seth. 2010. Medicinal (natureceuticals). Lice Sciences. plants as a source of antipyretic agents 78: 476-484. a review. Arch. Apll. Sci. Res. Naghibi, F.M., S. Mosaddegh, Mohammadi 2(3): 188-195. Motamed and A. Ghorbani. 2005. Singh, R.S., M. Uvarani and S.R. Raman. 2003. Labiatae family in folk medicine in Iran: Pharmacognostical and phytochemical From ethnobotany to pharmacology. studies on leaves of Anisomeles Iran J. Pharm Res. 4: 63-79. malabarica R. Br. AncSci. Life. 22(3): 106-110.

38 Journal of Agr. Research & Extension 30(3) (Suppl.): 29-39

Srinivasan, P., A. Sudha, P. Bharathajothi, Wiwanitkit, V. 2011. Concurrent malaria and K. Rajaguru, P. Rameshthangam, dengue infection: a brief summary and R. Manikandan and C. Arulvasu. comment. Asian Pac J Trop Biomed. 2010. Effects of anti-inflammatory and 1(4): 326-327. antipyretic activity of Anisomeles Vogel, A.I. 1978. Text book of Practical malabarica R.Br. Journal of Pharmacy Organic Chemistry. London: The Research. 3(7): 1598-1601. English Language Book Society and Sneader, W. 2005. Drug Discovery: a History. Longman. 368 p. Chichester. UK:Wiley. Zahir, A.A., A.A. Rahuman, C. Kamaraj, Tintinalli, J.E. 2010. Emergency Medicine: A. Bagavan, G. Elango, A. Sangaran A Comprehensive Study Guide and B.S. Kumar. 2009. Laboratory (Emergency Medicine (Tintinalli)). determination of efficacy of indigenous New York: McGraw-Hill Companies. plant extracts for parasites control. Parasitol Res. 105: 453-461.

39 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48

Isolation and Identification of Cyanobacteria from a Freshwater Aquaculture Pond in Northern Thailand

Dong Xia1,7, Norio Iwami2, Korntip Kannika3, Chayarat Pleumsumran4, Sirapran Fakrajang4, 6 4,5 1 Chayaporn Teercharernwong , Redel Gutierrez , Zhong Junsheng , Niwooti Whangchai4 and Tomoaki Itayama7* 1 Graduate School of Fisheries and Life Science, Shanghai Ocean University, Shanghai, China 2Faculty of Science and Engineering, Meisei University, Hino, Japan 3Phayao University, Phayao, Thailand 56000 4Faculty of Fisheries Technology and Aquatic Resources, Maejo University, Thailand 50290 5College of Arts and Sciences, Central Luzon State University, Science City of Munoz, Nueva Ecija, Philippines 3120 6Phayao Provincial Fisheries Office, Phayao, Thailand 56000 7 Graduate School of Engineering, Nagasaki University, Nagasaki, Japan *Corresponding author: [email protected]

Abstract

Harmful cyanobacteria such as Microcystis are frequently observed in freshwater aquaculture ponds. Due to their ubiquitous presence in these ponds, their characterization is very important from the standpoint of pond management. In this study two strains of cyanobacteria from a commercial tilapia/catfish pond in northern Thailand were isolated: a filamentous cyanobacteria and Microcystis. A PC- IGS gene locus of phycocyanin, which is a common gene in Cyanophyta, from the isolated strains was amplified to reveal the molecular base phylogenetic relationships. One was shown in the cluster of genus Microcystis in the phylogenetic tree. The other was placed in the cluster of genera Planktothrix and Oscillatoria. However, according to microscopic observation it was confirmed that the strain belonged to genus Oscillatoria due to the absence of gas vesicle in the cell.

Keywords: Microcystis, Oscillatoria, isolation, identification, PC-IGS, aquaculture pond

Introduction the health and welfare of humans and other animals because several species potentially produce Cyanobacteria are common members of cyanotoxin (Figueiredo et al., 2004). Microcystis the microscopic populations found in eutrophic water is especially a flagrant cyanobacterial genus systems (Carmichael, 1994). The extensive growth forming heavy blooms and often produces a strong of cyanobacteria, which is known as a blue-green hepatotoxin, microcystin. Blooms of Microcystis algal bloom, presents a considerable threat to spp. are not only frequently encountered in natural 40 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48 eutrophicated lakes and reservoirs but also size and concentrated to 30 mL. The concentrated in aquaculture ponds, since daily feeding for water sample was transferred to a 50-mL sterile the intensive aquaculture signifies the high plastic tube, kept in ice and transported to the nutrients loading to the ponds (Smith et al., laboratory within 24 hours. The sample was 2008). Many kinds of filamentous cyanobacteria immediately stored in a refrigerator upon arrival such as Oscillatoria spp., Planktothrix spp. and Anabaena spp., which potentially produce Isolation and cultivation microcystin and other cyanotoxins, are also The water sample was diluted to adequate frequently found in aquaculture ponds (Ruangrit density of cyanobacteria with the sterile M-11 et al., 2011). Such filamentous cyanobacteria as well culture medium (Hagiwara et al., 1984). Each as Microcystis spp. produce odorous metabolites, colony or filament of cyanobacteria in a droplet of which significantly compromise the economic water sample on a sterile glass slide was isolated value of the harvested fish (Whangchai et al., using a sterilized glass capillary under a 100x 2010). Hence, it is important for the aquaculture microscope. An isolated colony or filament was industry to consider these cyanobacterial-related transferred to a fresh sterile M-11 droplet on issues and its implications in aquaculture a new sterile glass slide for washing. The same production ponds. The adequate management of washing process was repeated three times. Each aquaculture pond water has been sought to cyanobacterial colony or filament was dropped control the proliferation of cyanobacteria. into a test tube filled with 3 mL M-11 medium. In this study, cyanobacteria from a The test tube was placed under a white fluorescent commercial aquaculture pond in northern Thailand light at 2,500~3,000 lux (12 hr light:12 hr dark) at were isolated and their molecular phylogenetic room temperature until blue-green color is visible analysis was performed using the sequence in in the medium which signifies the growth of a phycocyanin PC-IGS region (Neilan et al., 1995; cyanobacteria. Tillett et al., 2001). Extraction of DNA and PCR Materials and Methods The cyanobacterial cells used for DNA extraction were sonicated gently and harvested by Sample collection centrifugation at 14,000 x g for 15 min. Genomic A 10-L water sample containing cyanobacteria DNA was extracted using Fast DNA Spin Kit for was collected using a plastic bucket from Soil (MP Biomedicals, USA) in accordance with the surface of a private aquaculture pond, which the manufacturer’s instruction. The forward primer was mainly used for commercial farming of tilapia PCβF (5’-GGCTGCTTGTTTACGCGACA-3’) and the and catfish in northern Thailand. The water sample reverse primer PCαR (5’-CCAGTACCACCAGCAAC was filtered with a phytoplankton net, 20 µm mesh TAA-3’) were used because they are suitable for

41 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48 amplification of the PC-IGS region from a diverse DNA pellet at the bottom. After discarding the range of cyanobacteria (Neilan et al., 1995; Tillett supernatant, the DNA pellet was rinsed with 70% et al., 2001). The PCR was performed using ethanol using a vortex mixer and precipitated the following reaction mixture: 0.5 U of Taq again by centrifugation at 4oC, 15,000 rpm for 2 polymerase (Ex Taq HS, Takara Bio. Siga, Japan). min. After discarding the supernatant, the pellet The PCR required 30 cycles at 98oC for 10 sec, 55oC was dried up completely at room temperature. TE for 30 sec, and 72oC for 1 min. All PCR products buffer of 10 µL was added to the tube and mixed were analyzed by electrophoresis in 1.5% agarose well by vortex mixer. The purified PCR product in Tris-acetate-EDTA (TAE) buffer. was sent to a commercial sequencing service (Sigma Aldrich, Japan). The phylogenetic tree Sequencing and phylogenetic analysis was constructed from the determined sequences Each band of PCR products for the PC- from the PCR products of the isolated cyanobacteria IGS region was cut from the gel after electrophoresis. and the collected sequences of PC-IGS region Each gel fragment was transferred into a 1.5-mL from DDBJ using the free software MEGA (v5.2) plastic centrifugation tube with a pinhole at (Hall, 2011). the bottom (Nakayama and Nishikata, 1995). This tube was placed on the other 1.5 mL Results and Discussion plastic centrifugation tube. The small gel fragments were collected through the pinhole after centrifugation Isolation of Cyanobacteria from an Aquaculture at 8,000 rpm for 5 min. Adequate volume of Pond equilibrated phenol was added to the tube and The major cyanobacterial species in the then mixed well in a vortex mixer. The gel in sample were clearly identified as Microcystis spp. the tube was frozen in liquid nitrogen for 30 min. and filamentous cyanobacteria such as Planktothrix After melting at room temperature, the tube was spp. and/or Oscillatoria spp. by their remarkable centrifuged at 15,000 rpm for 5 min. The top water morphology (Komárek and Hauer, 2013). The two layer containing DNA was transferred to a new species of cyanobacteria were obtained by 1.5 mL centrifugation tube. After mixing it with the isolation method previously described. No equal volumes of PCI (Phenol: Chloroform: other species was observed for half a year after Isoamyl Alcohol, 25:24:1) using a vortex mixer, the successive subcultures of the two isolated the tube was centrifuged again (15,000 rpm, species in M-11 culture medium. However, many 5 min). The water layer was transferred to a new bacteria could have coexisted because the isolations 1.5 mL tube. Equal volumes of isopropyl alcohol were performed under non-axenic condition. and 1/10 volume of 3 M sodium acetate were Optical micrographs of the cyanobacteria added to the tube. Then, the tube was centrifuged are presented in Figure 1. They were temporarily again (4oC, 15,000 rpm, 20 min) to precipitate as MC-FPNT because they were morphologically

42 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48 identified under the genus Microcystis (Komárek cell were observed, it could be identified as genus and Hauer, 2013; Komárek and Cyanobacterial, Planktothrix (Komárek and Komárková, 2004). 2006; Crow, 1923). The diameter of the cell was Therefore, it was temporarily named OC-FPNT. 4~5 µm. The colony (Figure 1) consisted of The cell diameter in a trichome was 2~3 µm. several sub-colonies in which several hundreds of The cell length was 4~8 µm. However, it is cells were loosely packed. The clear mucilaginous well known that morphological identification envelope (sheath) wrapped each colony. Figure of cyanobacteria is limited in the phylogenetic 1(B) was the cyanobacterial genus Oscillatoria analysis (Shigeto et al., 2000). Therefore, molecular due to the absence of gas vesicles (Komárek and phylogenetic method had to be applied to establish Komárková, 2004). If several gas vesicles in the the identities of the cyanobacteria.

Figure 1 Optical micrographs of isolated cyanobacterial colonies and filaments. The colonies which have morphology of Microcystis spp. are shown in (A). A thin sheath surrounds the colonies. Solid arrows in (A) show several positions where the sheath is clearly visible. The cyanobacterial filaments are shown in (B). Individual cells are visible in a trichome. The solid arrow in (B) shows the cell which is clearly visible in a trichome.

43 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48

Figure 2 PCR products from the cyanobacterial PC gene related to genus-specific sequence signatures M-DNA ladder marker; A-MC-FPNT; B-OC-FPNT; C-Microcystis aeruginosa NIES843

. Cyanobacterium-specific PCR other hand, the DNA bands for OC-FPNT Oligonucleotide primers (PCβF and appeared above the bands of NIES843 and MC- PCαR) specific to the conserved regions of the FPNT. The DNA sequence of each PCR product cpcB and cpcA genes encoding the β- and α- was determined after the extraction and the phycocyanin subunits of phycobilisomes of purification from the agar gel. cyanobacteria were used to amplify a DNA fragment containing the intervening intergenic Phylogenetic analysis spacer region (cpcBA-IGS) (Neilan et al., 1995; Ten sequences of Microcystis spp., 6 Tillett et al., 2001). cpcBA-IGS is a useful region sequences of Planktothirix spp., 4 sequences of for cyanobacterial DNA samples from non-axenic Oscillatoria spp. and 3 sequences of Anabaena culture because phycocyanin is a cyanobacterial spp. were collected from the DNA Data Bank specific chromoprotein in the photosynthetic of Japan (DDBJ). Each accession number in the system. The amplicon size of PCR for the region GeneBank was described in the legend of Figure is 500-740bp with the majority of strains providing 3. Multiple alignment for the collected sequences about 700bp products (Neilan et al., 1995).Figure and the two sequences of the isolated cyanobacteria 2 shows the results of PCR products. The DNA were performed by MUSCLE in MEGA (v5.2) before bands for Microcystis aeruginosa NIES843 and constructing the phylogenetic tree by the neighbor- MC-FPNT were shown at the position of around joining algorithm (Tajima-Nei model) with 700bp. The true size of the PCR product is 663bp the bootstrap test of 500 times using MEGA (v5.2) according to the whole genome sequence data of (Hall, 2011). NIES843 (Accession Number: AP009552). On the

44 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48

Figure 3 shows the estimated tree. Each including genera Planktothrix, Oscillatoria and branch shows more than 50% reliability as Anabaena. Cluster I consists of two major branches percentage of the bootstrap. There is no root I-A and I-B. Cluster I-A includes Microcystis in the tree, because it was difficult to add aerugisnosa NIES843 isolated from a Japanese an adequate sequence of PC-IGS distant from Lake. The other branch I-B includes the isolated such cyanobacteria in the evolution as an outer Microcystis type cyanobacteria (MC-FPNT), which OTU. The phonetic relationships inferred revealed is nearest to Microcystis protocystis SPC697 and an essentially bifurcating phylogeny, with one Microcystis spp. M124CC01. Microcystis protocystis branch (I) dominated by genus Microcystis and SPC697 was isolated in Brazil according to the the others (II) consisting of filamentous cyanobacteria annotation of the sequence (FJ801045).

A

I

B

A

II B

C

Figure 3 PC-IGS region DNA distance tree. Included in the alignment were the Microcystis spp., Planktothrix spp., Anabeana spp., Oscillatoria spp. and Geitlerinema spp. PC-IGS sequences obtained from the DDBJ database. The phylogenetic tree was reconstructed using the neighbor-joining algorithm as implemented by MEGA (v5.2). The scale bar shows the genetic distance. The number associated at each branch shows the reliability in percentage of bootstrap (500 times).

45 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48

The morphology of Microcystis protocystis Conclusion was described as the colonial envelope quite indefinite, cells were very loosely scattered Two isolates of cyanobacteria were according to Crow (1923). The cells of MC-FPNT obtained from a commercial tilapia/catfish pond in were loosely scattered, but the envelope appeared north Thailand. One was first estimated to be of clear (Figure 1A). It is in fact well known that Microcystis sp. by morphological screening. The morphological aspect does not correspond to molecular base identification using the PC-IGS genetic distance (Shigeto et al., 2000). region confirmed it to be of genus Microcystis. It Cluster II was divided into three branches, is especially nearest to Microcystis protocystis in designated as II-A, II-B, and II-C. Clusters II-A and the phylogenic tree. The other was placed in the II-C represented the population of the filamentous cluster of genera Planktothrix and Oscillatoria in genera Planktothrix and Oscillatoria. Cluster II-B the tree. Moreover, according to the microscopic was a branch of genus Anabaena. The molecular observation, it contained no gas vesicle in the base identification using PC-IGS represented that cell, which is a significant feature of genus OC-FPNT was very close to Oscillatoria spp. Oscillatoria. Therefore, it was identified as CYA469. This result agreed with the morphological Oscillatoria sp., which was nearest to Oscillatoria identification that OC-FPNT was in the genus sp. CYA469 in the phylogenic tree. The two Oscillatoria (Komárek and Komárková, 2004). identified cyanobacterial isolates must play an Both isolates were successfully placed in active role in basic eco-physiological studies and the phylogenetic tree based on the sequence of further investigations on the adequate management the PC-IGS region. However, the available number of cyanobacteria in aquaculture ponds. of sequences of PC-IGS in the GeneBank was limited in comparison with the number of available Acknowledgements sequences of 16SrDNA, a standard DNA region in bacterial systematics (Marchesi et al., 1998). This research was supported by the Ministry Therefore, it is desirable to obtain the full sequence of Education, Science, Sports and Culture, Japan, data of 16SrDNA from the axenic cyanobacteria Grant-in-Aid for Scientific Research, (B) 21404012 to identify the exact position in the phylogenic tree more widely covered with cyanobacterial species and genera.

46 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48

Reference Marchesi, J.R., T. Sato, A.J. Weightman, T.A. Martin, J.C. Fry, S.J. Hiom, and W.G. Wade. Carmichael, W.W. 1994. The toxins of 1998. Design and evaluation of useful cyanobacteria. Sci. Am. 270: 78-86. bacterium-specific PCR primers that Crow W.B. 1923. The Taxonomy and Variation amplify genes coding for bacterial 16S of the Genus Microcystis in Ceylon. rRNA. Appl. Environ. Microbiol. New Phytologist. 22(2): 59-68. 64: 795-799. Figueiredo, D.R., U.M. Azeiteiro, S.M. Esteves, Nakayama, H. and T. Nishikata. 1995. Bio- F.J. Goncalves and M.J. Pereira. 2004. Experiment illustrated 2 Basics of Microcystin-producing blooms-a serious Genome Analysis. Tokyo: SHUJUNSHA. global public health issue. Ecotoxicol. Neilan, B.A., D. Jacobs, and A.E. Goodman. 1995. Environ. Saf. 59: 151-163. Genetic Diversity and Phylogeny of Toxic Hagiwara, T, O. Yagi, Y. Takamura and R. Sudo. Cyanobacteria Determined by DNA 1984. Isolation of bacteria-free Polymorphisms within the Phycocyanin Microcystis aeruginosa from Lake Locus. Appl. Enviro. Microbiol. Kasumigaura. Jpn. J. Water Poll. Res. 61: 3875-3883. 7: 437-442. Ruangrit, K., N. Whangchai, J. Pekkoh, Hall, B. G. 2011. Phylogenetic Trees Made W. Ruangyuttikarn and Y. Peerapornpisal. Easy: A How-to Manual 4thEd. 2011. First report on microcystins Sunderland: Sinauer Association. pp. 4-6. contamination ingiant freshwater prawn Komárek, J. and Cyanobacterial Taxonomy. 2006. (Macrobrachium rosenbergii) and Nile Current Problems and Prospects for the tilapia (Tilapia nilotica) cultured in earthen Integration of Traditional and Molecular ponds. Int. J. Agric. Biol. 13: 1025-1028. Approaches. Algae. 21(4): 349-375. Shigeto, O., S. Shoichiro, L. Renhui, M. Satoshi Komárek, J. and J. Komárková. 2004. and W.M. Makoto. 2000. Morphological Taxonomic review of the cyanoprokaryotic variability of colonies of Microcystis genera Planktothrix and Planktothricoides. morphospecies in culture. J. Gen. Appl. Czech Phycology. 4: 1-18. Microbiol. 46: 39-50. Komárek, J. and T. Hauer. 2013. CyanoDB.cz - Smith, J.L. G.L. Boyer. and P.V. Zimba. 2008. On-line database of cyanobacterial A review of cyanobacterial odorous and genera. - Word-wide electronic bioactive metabolites: Impacts and publication. [Online]. Available management alternatives in aquaculture. http://www.cyanodb.cz. (14 May 2013) Aquaculture. 280: 5-20.

47 Journal of Agr. Research & Extension 30(3) (Suppl.): 40-48

Tillett, D., D.L. Parker, and B.A. Neilan. 2001. Whangchai, N., C. Pleumsumran, S. Fakrajang, Detection of toxigenicity by a probe for N. Iwami and T. Itayama. 2010. Musty the microcystin synthetase A gene (mcyA) Odor in Tilapia (Oreochromis niloticus) of the cyanobacterial genus Microcystis: Cultured in Cages and Earthen Ponds. comparison of toxicities with 16S rRNA J. Agr. Res. & Ext. 27(1): 19-27. andphycocyanin operon (Phycocyanin Intergenic Spacer) phylogenies. Appl. Environ. Microbiol. 67: 2810-2818.

48 Journal of Agr. Research & Extension 30(3) (Suppl.): 49-55

Carbon Footprint of Central Canteen of Mahidol University Salaya Campus, Thailand

Sayam Aroonsrimorakot*, Chumporn Yuwaree, Chumlong Arunlertaree Rungjarus Hutajareorn and Tarinee Buadit Faculty of Environment and Resource Studies, Mahidol University, Nakhonpathom, Thailand 73170 *Corresponding author: [email protected]

Abstract

All human activities generate greenhouse gases (GHG) which in turn cause the greenhouse effect, and therefore harm the earth. The concept of measuring GHG generated from activities in an organization can be helpful in reducing and managing GHG’s. This study is aimed at measuring and calculating the GHG emissions or carbon footprint of the central canteen activities at Mahidol University, Salaya campus in the year 2010 (January-December) in units of Carbon Dioxide equivalent (CO2e). Data was collected from various sources of GHG such as electricity, water, LPG and fuel utilization including wastewater and waste and then multiplying it with the Thai standard impact factor value. The result found that the volume of GHG emission was 489.16 tons CO2e, which means that the average GHG emission per shop was 13.22 tons CO2e and average GHG emission per head of student was 0.02 tons CO2e. The significant sources of GHG were electricity utilization and waste generation, respectively. Appropriate technology will be used in the future to reduce significant resource use and manage waste generation.

Keywords: carbon footprint, greenhouse gases, central canteen of Mahidol University

Introduction Protocol; carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), hydrofluoride perfluoro Carbon footprinting is used for companies carbons (HFCs), perfluoro carbons (PFCs) and to assess the greenhouse gas (GHG) emissions sulfur hexafluoride (SF6). resulting from their activities, both directly and Carbon footprint can be calculated by indirectly, such as the burning of fuel, electricity using the life cycle assessment (LCA), which is the consumption, and waste management and international standard ISO 14040, 14044, used transport, by showing the amount of greenhouse for the assessment of environmental impact gas emissions in units of carbon dioxide equivalent throughout the life cycle. It can be calculated

(CO2e). Six species of greenhouse gas are used from the formula: CO2 equivalent of each process to assess the carbon footprint under the Kyoto = Amount of activity x CO2 emission intensity.

49 Journal of Agr. Research & Extension 30(3) (Suppl.): 49-55

Total amount of every type of greenhouse gas from Materials and Methods all activities, which are then converted into units of carbon dioxide equivalents by multiplying total Setting organization boundaries emissions of each type of greenhouse gas with its This study covered the central canteen of global warming potential (GWP), this then gives Mahidol University, Salaya campus and consists of the Carbon Footprint of the organization. one building with 37 stores as shown in Figure 1.

Figure 1 Central canteen of Mahidol University, Salaya campus

Setting operational boundaries Data inventory Identify all sources of greenhouse gas The data collected in this study were from emissions both direct and indirect sources, which relevant documents of both primary data (solid can be divided into the 3 scopes as follows: waste) and secondary data (electricity and water Scope 1: Amount of greenhouse gases supply, quantity and quality of the wastewater, generated by sewage treatment processes. and the use of LPG). Scope 2: Emissions arising from energy For solid waste arising, the researchers imports such as liquefied petroleum gas (LPG) took into account only food waste because it is the consumption and the purchase of electricity and mostly produced type of waste from the canteen. (as water supply. shown in Figure 2) Scope 3: Indirect greenhouse gases emissions produced by generated waste.

50 Journal of Agr. Research & Extension 30(3) (Suppl.): 49-55

Figure 2 Weighing of food waste in the central canteen of Mahidol University

Calculation of GHG emission Climate Change (IPCC) or from a national database GHGs can be calculated by multiplying of each country, and so on. The example of the emission factor, which is commonly used emission factors that are used in the study are internationally, and in accordance with the shown in Table 1. guidelines of the Intergovernmental Panel on

Table 1 The example of emission factors that are used in this study. (Thailand Greenhouse Gas Management Organization, 2011)

Sources of GHGs Unit roFaaFmnoissimE References

(kg CO2) Water supply m3 0.0264 Metropolitan Waterworks Authority (Thailand) Electricity kWh 0.5610 TC Common data LPG from natural gas kg 0894.0 Thai LCI data

Results and Discussion University, Salaya campus in the year 2010 and the emission factors used to calculate the amount Resource consumption and food waste of each greenhouse gas sources are presented in generated by the central canteen at Mahidol Table 2.

51 Journal of Agr. Research & Extension 30(3) (Suppl.): 49-55

Table 2 Consumption of resources and the amount of waste generated

GHGs Sources Resource consumption/ Emission factor Unit Wasteogenerated Scope 1:

Quantity of wastewater 7,89.,.00 L CH4 Emission factor = 083 kgCH4 N2O Emission factor = 0.005 kgN2O-N Scope 2: Tap water used 4,185 m3 0.0264 m3 Electricity consumption 924,59.81 kWh 0.5160 kWh LPG consumption 10,758 kg 0.4980 kg Scope 3: Food waste generated 72,043.94 kg 2.5300 kg

The total greenhouse gas emissions that 2010 are equal to 489.16 tons CO2e as shown in resulted from the activities of the central canteen Figure 1, and the ratio of the amount of greenhouse of Mahidol University, Salaya campus in the year gases generated by each activity is seen in Figure 2.

Total of GHGs Emissions (tons CO2e)

200 180 160 140 120 100 80 60 40 20 0 Wastewater Electricity Water supply LPG consumption Food waste treatment consumption generated

Figure 3 The amount of greenhouse gases resulting from the activities in the central canteen of Mahidol University, Salaya campus in the year 2010

52 Journal of Agr. Research & Extension 30(3) (Suppl.): 49-55

Total of GHGs Emissions (tons CO2e)

18% Wastewater treatment 37% Electricity consumption Water supply LPG consumption 38% Food waste generated 7%

0% Figure 4 Percentage of greenhouse gases emissions of each activity

In summary, in 2010, the central canteen of greenhouse gases by the central canteen of of Mahidol University, Salaya campus released Mahidol University in each scope is shown in greenhouse gas emissions from various activities in Table 3. the amounting to 489.16 tons CO2e. The emission

Table 3 The amount of greenhouse gases that resulted from the activities in the central canteen of Mahidol University

Greenhouse gases source Total of GHGs Emissionso

(tons CO2e) Scope 1: Direct greenhouse gas emissions  Greenhouse gases generated by the wastewater treatment 86.98 Scope 2: GHGs emissions arising from the importation of energy  Electricity consumption 184.65  Water supply 0.25  LPG used 35.01 Scope 3: Indirect greenhouse gas emissions  Solid waste generated 182.27

53 Journal of Agr. Research & Extension 30(3) (Suppl.): 49-55

Conclusion the source has a higher emission factor used to calculate the amount of greenhouse gas emissions, The total of the greenhouse gas emissions the results obtained may be much higher than by the central canteen of Mahidol University, a greater source with a lower emission factor.

Salaya campus is 489.16 tons CO2e. The source The time of the academic calendar is another of greenhouse gases which has the highest factor that affects the amount of greenhouse emission is electricity consumption, followed by gases. This is due to the fact that the in-class the solid waste generated. The source with the semester periods and vacation periods have minimal release is water supply. different levels and types of student and staff The average GHG emission per shop activities. was 13.22 tons CO2e and average GHG emission per head of student was 0.02 tons CO2e. Alternatives to compensate for the greenhouse The canteen is a major source of gas emissions of the organization: greenhouse gases is the use of electricity such Planting trees to absorb carbon dioxide: Teak as lamps, fans and TV including the use of tap plantations can produce carbon sequestration water for cooking food. Cleaning and washing with the approximate amount of 4.61 tons/rai/year dishes result in wastewater that is a direct source (1 rai = 1,600 SqMts). For northern black wattle, of the greenhouse gases. In addition, cooking the amount of carbon sequestration is about 9.08 and eating contribute to waste as well. In this tons/ha/year. As such, if the central canteen of study, the collection of waste in the canteen was Mahidol University, Salaya campus is to reduce only based on food waste since it is the major the amount of carbon dioxide, it should 106.11 rai generated of waste types and is produced in of teak or 53.87 rai of northern black wattle. large quantities. In addition to electricity and water consumption, another source of greenhouse gases Carbon credit (CERs): It can be observed that comes from energy imports. The canteen has through the three cooperative mechanisms of the another major source of energy consumption in the Kyoto Protocol, carbon credits become more like form of LPG for cooking this is widely used in a commodities, and they are tradable or exchangeable variety of restaurants. However, there is only in a market known as the “carbon market.” If the a minor amount of greenhouse gas emissions from central canteen activities of Mahidol University, electricity usage and garbage production. Salaya campus wants to reduce the amount of

The amount of greenhouse gas emissions GHGs by 489.16 tons CO2e per year by purchasing is not only determined by the amount of resources CERs at the price of the ECX Dec '11 market, the used or the amount of waste that is generated faculty will have to pay the total cost of 270,675852 occurs. It also depends on the emission factor. Baht or will pay a total cost of 271,312840 Baht for Although the source amount may be less, if the BlueNext Spot market. Carbon credit trading

54 Journal of Agr. Research & Extension 30(3) (Suppl.): 49-55 market at 21-25 March 2011 from Thailand Reference Greenhouse Gas Management Organization. (Thailand Greenhouse Gas Management Intergovernmental Panel on Climate Change. Organization, 2011) 2007. Wastewater Treatment and Discharge. [Online]. Available http:// Constraints www.ipcc-nggip.iges.or.jp/public/2006 Wastewater quality data used to calculate gl/vol5.Html (2011 June 13). the amount of greenhouse gas emissions is the Thailand Greenhouse Gas Management data of year 2009 due to data limitations of Organization. 2011. Emission Factor. monitoring wastewater quality of the university. [Online]. Available http://thaicarbonlabel. Amount of wastewater generated cannot tgo.or.th/ download/Emission_Factor.pdf be calculated correctly because of a lack of a (2011 April 20). proper measuring system. Therefore, this study Thailand Greenhouse Gas Management calculates the amount of wastewater from 80 Organization. 2011. Carbon Market. percent of the amount of water used. Obtained [Online]. Available http://www.tgo.or.th/ values have error. index.php?option=com_content&task=sec Data collection of LPG usage by restaurants tion &id=4& Item id=810 (2011 May 10). in the canteen is an inquiry from the caterer. It is Thailand Greenhouse Gas Management an estimated amount of LPG used each month, Organization. 2011. GHG & Climate and not the exact value. Change. [Online]. Available http:// www. tgo.or.th/index.php?option=com_content& task=section &id=1&Item id=2.. (2011 January 20). United Nation Framework Convention on Climate Change. 2011. United Nation Framework Convention on Climate Change. [Online]. Available http://unfccc.int/2860.php (2011 January 20).

55 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Mathematical Model of Freeze Drying on Mango

Sakawduan Kaewdam*, Chanawat Nitatwichit, Jatupong Varith and Somkiat Jaturonglumlert Division of Food Engineering, Faculty of Engineering and Agro-Industry, Maejo University, Chiang Mai, Thailand 50290 *Corresponding author: [email protected]

Abstract

The mathematical model of freeze-drying on mango was examined. The mango cultivar ‘Nam Dok Mai’ was frozen at -40oC and freeze-dried at -40oC for 6 hours then at -20oC 10 hours and then at -10oC 6 hours, The temperature and time of secondary drying were 10, 20, 30oC and 2, 4, 6, 8 hours respectively, The pressure during the drying process were 20, 30, 40 and 80 Pa. It was shown that the optimal conditions were in the secondary drying stage with at 10oC for 6 hours and 20 Pa. After drying, the quality of mangoes was determined. The hardness was 6.1834 N and the water activity was 0.276. The colors L* a* and b* value were 79.86, 4.29 and 53.62 respectively. The final moisture content of the product was 6.8% and the specific energy consumption was 253.07 kWh/kg. The average drying rate was

0.294 gH2O/g dry mass-h and the effective moisture diffusivity coefficient from linear equation method ranged from 5.54  10-11to 2.90  10-10 m2/s. Thin-layer drying models of Newton, Page , Modified Page and Henderson and Pabis were evaluated based on coefficient of determination (R2), reduced chi-square (  2 ) and Root Means Square Error (RMSE). The Modified Page model was found to be the better model, with coefficient of determination of 0.998, reduced chi-square value of 0.00026 and RMSE at 0.016215.

Keywords: mathematical model, freeze drying, mango

Introduction supplementation. It was suggest to that addition of mango to generally accepted healthy diet could have Mango (Mangifera indica L.) is a prominent a beneficial effect in preventing hypertriglyceridemia tropical fruit which is widely grown in Thailand. (Robles et al., 2011). There is an increasing trend The fruit is rich in antioxidants and recommended of consumption health food by Thai and foreigners. to be included in the daily diet due to its health However, there is harvest limit on materials. benefits such as reduced risk of cardiac disease, For example, mangoes can be harvested during anti-cancer and anti-viral activities (Sivakumar et al., May to June only. Freeze drying technology is used 2011). Recent epidemiological study also stated in the manufacture of pharmaceutical products, that the consumption of mangoes significantly herbal and food, supplements and the storage of reduced serum triglycerides after 03 days of microorganisms. In food processing, freeze drying

56 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67 is used to preserve food. Most of the initial raw Materials and Methods material properties such as shape, appearance, taste, color, flavor, texture, biological activity etc Preparation: were preserved. Freeze-dried fruits are used in Fresh ripe mangoes (cultivar ‘Nam Dok some breakfast cereal or sold as snack. They are Mai’) were selected for their homogeneity in size, popular snacks of choice among toddlers, weight, peel color, and density by floatation in 4– preschoolers and dieters, as well as being used 5% NaCl solution. The total soluble solid content by some pet owners as a treat for pet birds. of mangoes was determined to be in the range of Culinary herbs are also freeze-dried, although air- 16–20°Brix (Nathdanai et al., 2011).The mangoes dried herbs are far more common and less were peeled, cut into 3 cm x 3 cm x 1 cm by mold expensive. Freeze drying is an alternative for food and frozen at -40oC as shown in Figure 1. They processing to meet the needs of consumers. were there freeze-dried in a freeze dryer (Freeze Freeze dried food is a large industry. Due dryer Heto Powerdry PL3000). to high production cost and the need for freeze The freeze dryer accessories which include dryers from abroad, the small and medium a vacuum pump, condenser, chamber and heating enterprises (SME) could not invest. This research plate. Initially a vacuum pump reduces the ambient aims to study the mathematical model for freeze gas pressure in an external chamber to a vacuum drying mango as a guide line to scale up the state. The sample is heated to vaporize the manufacturing in SME and to determine the costs sublimation. Then a vacuum pump sucks the for investors steam from the chamber which condenses into ice in a condenser as shown in Figure 2.

Figure 1 Mold (a) and sample (b)

57 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Figure 2 Diagram of Freeze drying equipment.

Experimental design: Quality analysis: The experimental design is show in Table Moisture content: Dried mango samples of 5 g 1. Primary drying was separated into three sets. were placed in Infrared moisture determination o o Primary drying 1 (Tf1) -40 C for 6 hours (tf1). balance (AND AD-4714A, Tokyo, Japan) at 105 C o Primary drying 2 (Tf2) -20 C for 10 hours (tf2) and for 90 minutes. Three samples from each trial o Primary drying 3 (Tf3) -10 C for 6 hours (tf3) were used for the moisture determination and the (Suchada et al., 2546). Secondary drying temperature average moisture content was reported. o (Ts) was 03, 20 and 30 C, respectively for 2 hours. Hardness: The hardness was determined by a Total drying time was 24 hours and the drying texture analyzer (Texture Analyzer TA.XT2i.plus, pressure was 20 Pa in all experiments. (Run 1-3) UK). A cylinder probe (2 mm diameter) was used From the optimal secondary drying temperature, for puncture compression analysis. The probe was the secondary drying time (ts) was 4, 6 and 8 hours used to measure the maximum force required to and total drying time was 26, 28 and 30 hours penetrate an individual dehydrated piece of mango, respectively. The pressure of drying was 20 Pa in positioned horizontally over a 9.3 mm diameter all experiments. (Run 4-6). From the optimal hole. The measurement settings on the Texture secondary drying temperature and time, the pressure Analyzer were pre-test speed of 3.0 mm/s and of drying (P) was 40 60 and 80 Pa, respectively. test speed of 5.0 mm/s, hardness was measure, The moisture content (MC), hardness (H) water Three replicates for each treatment were performed. activity (aw) and color of the finished products were measure.

58 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Table 1 Experimental design

Primary drying 1 Primary drying 2 Primary drying Secondary Pressur ttotal RUN o o o o Tf1 ( C) tf1 (hr) Tf2 ( C) tf2 (hr) Tf3 ( C) 3 tf3 (hr) Ts ( dryingC) ts (hr) Pe (Pa)ee (hr) 1 -40 6 -20 10 -10 6 10 2 20 24 2 -40 6 -20 10 -10 6 20 2 20 24 3 -40 6 -20 10 -10 6 30 2 20 24 4 -40 6 -20 10 -10 6 30 4 20 26 5 -40 6 -20 10 -10 6 30 6 20 28 6 -40 6 -20 10 -10 6 30 8 20 30 7 -40 6 -20 10 -10 6 30 6 40 28 8 -40 6 -20 10 -10 6 30 6 60 28 9 -40 6 -20 10 -10 6 30 6 80 28

2 2 2 Water activity: Water activity was measured were E  (L*L0 )  (a*a0 )  (b*b0) (1) performed using water activity meter (AQUA Lab 3TE, USA). Samples were placed in the appropriate Structures: The internal structures of freeze-dried plates and measured. Three samples from each mangoes were characterized with a scanning trial were used and the average water activity is electron microscope (SEM) (JEOL JSM-5410V, reported. Tokyo, Japan). Cross-sectioned samples were affixed Color: Color values L*, a* and b* were measured on silver-painted sample holders and covered with using Spectrophotometer (HunterLap MiniScan a fine layer of gold in a sputter coater (PSI sputter XE plus, Germany). To obtain representative color coater, PA, USA) under vacuum. The coated of the samples, the dried samples were ground to samples were photographed with SEM at 10 kV powder using a small‐scale blender to obtain representative colors. A 1 g sample of mango Effective moisture diffusivity coefficient, Deff powder was put in a 5 cm plastic Petri dish. The Considering unidimensional moisture lens of the colorimeter, covered with plastic wrap movement and assuming that the mango sample was placed directly on the mango powder to was a continuous infinite rectangular slab (Wang measure the color values. Three measurements and Brennan, 1995) the effective mass diffusivity were made of each sample, and the average was estimated by using Fick’s diffusion model value was reported. The L* color value indicates the degree of brightness or whiteness of the product. M 2M  D (2) The a* and b* color values indicates the degree of t z2 redness and yellowness, respectively. The Total Color Difference ( E ) was determined by Equation 1. 59 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Moreover, assuming uniform initial moisture an energy intensive operation of some industrial distribution, negligible external resistance and significance (Sharma and Prasad, 2006). The SEC isothermal process, the solution proposed by was estimated and expressed following Equation (Crank, 1975). is: 6 when the enthalpy of drying air was estimated from its psychometric properties.

81 (2n1)(D)t22 MRexpeff (3) Eheater  Evacuum  Ecooling 22  SEC  (6) n0 (2n1)L  M w

For sufficiently long drying times, using When Eheater is energy required to heat the first term in the series is adequate, Equation 4 the air, )kWh) Evacuum is energy requirement of (Doymaz, 2007). the vacuum pump. )kWh) Ecooling is energy requirement of the compressor )kWh) and MW is the amount of moisture removed during the drying 8  2 (D )t  MR  exp eff  (4) 2  2  process )kg)   L 

Drying model of freeze-drying The diffusion coefficient could be calculated Data obtained from the measurements of by plotting experimental drying data in term of Ln moisture content were expressed as a percentage (MR) versus time. The effective moisture diffusivity wet basis and then converted to gram water per is the slope of the linear segment, Equation 5 gram dry matter. The experimental drying data for (Tunde and Ogunlakin, 2011). dried sample were fitted to the exponential model thin layer drying models. Six models, Newton, Page, Modified Page, Henderson-Pabis, Logarithmic and slopeL2 Wang and Sing were selected to describe the freeze D   (5) eff 2 drying process as shown in Table 2 They were selected because of their simplicity, high correlation to most drying data and common use in the literature Specific energy consumption, SEC by using non-linear regression analysis (Dayang The drying of food material, a process of et al., 2012) simultaneous heat and mass transfer, represents

60 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Table 2 Thin layer model

Model name Model expression Newton MR=exp(-kt) Page MR=exp(-ktn) Modified Page MR=exp((-kt)n) Henderson and Pabis MR=aexp(-kt) Logarithumic MR=a exp(-kt) + c Wang and Sing MR=1 + at + bt2

The Moisture Ratio (MR) can be calculated Boer model (G.A.B.) because It is most widely from Equation 7 used for food with high sugar content such as mango (Rangel et al., 2011). following Equation 8

M  M MR  eq (7) Min  Meq A BCa M  w (8) eq 1 C1 Ca  BCa  w w

When Meq is Equilibrium moisture content and Min is Initial moisture content When aw is Water activity and A, B and C The equilibrium moisture content can be is constant as shown in Table 3 calculated from the Guggenheim-Anderson-de

Table 3 The constant for G.A.B. equation

Parameter Temperature (K) 288.15 298.15 308.15 A 0.1686 0.1242 0.1175 B 0.9376 0.9399 0.9363 C 15.0253 15.9943 15.1964

The coefficient of determination (R2) was the relative goodness of fit. The best model one of the primary criteria to select the best model describing the drying behavior of freeze drying to compare with the experimental data. In addition was chosen as the one with the highest R2 and to R2, reduced chi-square (χ2) and Root Means the least RMSE (Othman et al., 2012). This Square Error (RMSE) were also used to compare parameter can be calculated following Table 4.

61 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Results and Discussions purchased from supermarket and properties were examined. The Total Color Difference ( E ) Qualities and optimum conditions between experimental and reference product is It was found that, each experiment did not also presented. differ and there was shrinkage in some experiments. It can be observed that, RUN 5 had the The color of the product changed very little compared lowest value (4.37). From the final moisture to the samples before drying as shown in Figure 3. content and qualities of the dried mango it was Qualities of the dried mangoes are shown in concluded that. RUN 5 is the best conditions for Table 4. The moisture content of the sample from freeze drying of mango. It was the secondary different drying process as RUN 5 gave the lowest drying stage with in the temperature at 10oC for 6 moisture content (6.80%) whereas RUN 1 gave hours and the pressure of 20 Pa. This was a condition the highest value of 11.90%. Water activity of that led to a mathematical modeling study. sample corresponded to the moisture content. The relationship between drying rate (gH2O/ The dried sample had aw of 0.23-0.34, which were gdry mass-h) and moisture content( d.b.) is shown in within the recommended level for safe storage (0.6). Figure 4. The drying rate was falling and was 0.19-

The hardness of freeze dried sample was their 0.38 gH2O/gdry mass-h. with an average of 0.294 gH2O/ highest at 8.2872 N. Reference product were gdry mass-h.

Table 4 Qualities of the dried mangoes

RUN MC H aw Color (%) (N) L* a* b* E 1 11.90 2.8166 0.266 79.30 3.56 45.69 7.28 2 10.80 5.9069 0.326 77.35 6.86 51.87 13.94 3 9.20 7.6867 0.238 78.72 5.51 50.09 11.84 4 8.50 3.2798 0.230 77.61 3.43 49.14 10.80 5 6.80 6.1834 0.276 79.86 4.29 42.62 4.37 6 7.50 6.6459 0.340 81.12 2.70 43.03 5.18 7 7.50 6.7275 0.336 81.57 3.24 42.68 5.03 8 9.20 7.8953 0.334 80.71 4.86 46.66 8.54 9 8.50 8.2872 0.301 80.84 3.01 42.89 4.89 Ref. Product 6.60 5.3971 0.230 78.92 3.53 38.42 -

62 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

0.40 h) - 0.35

0.30 dry mass dry

/(g 0.25

H2O 0.20

0.15

0.10

Drying rate,g 0.05

0.00 0.07 0.08 0.31 0.37 0.88 2.28 3.98 Moisture content (d.b.)

Figure 4 Relationship between drying rate (gH2O/gdry mass-h) and noisture content (d.b.)

The microstructure of a cross section of through plasmalemma membrane boundaries. mango freeze dried under different drying processes The most important pathway for water movement was investigated by SEM. During drying, water in through plant tissues is through the cell wall. as the mangoes was transported through several shown in Figure 5. Cross section of freeze dried possible pathways (Tyree, 1970). In the first pathway, mangoes from optimal condition figure showing water passes from one cell to the next via small porous structure. While, cross section of cytoplasmic strands (plasmodesmata). In the second, freeze dried mangoes from non optimal condition. pathway water alternately enters and leaves Figure had a big hole in the center of the dried successive cells along its pathway by passing mango.

(a) (b) Figure 5 Optimal condition (a) and non optimal condition (b)

63 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

4E-10 3.5E-10

3E-10 /s

2 2.5E-10 , m ,

eff 2E-10 D 1.5E-10 1E-10 5E-11 0 0.07 0.08 0.31 0.37 0.88 2.28 3.98 Moisture Content , d.b.

Figure 6 Relationship between effective moisture diffusivity coefficient and moisture content

Effective moisture diffusivity coefficient, Deff the energy used for heating is less than the amount Considering the slope of the graph of relation of energy. In RUN 4 , 5 and 6 shows, The SEC is of experimental drying data in term of Ln(MR) directly proportional to time, more time for drying. versus time, plot Deff versus moisture content of The SEC increase. RUN 6 gave the highest SEC mango freeze dried. (Figure 6) The Deff increases of 277.12 kWh/kg And RUN 7, 8 and 9 increase as moisture content increases and became almost in pressure reduced energy consumption because constant at low moisture content. Deff was in the the vacuum pump performed less work. range of 5.54  10-11 to 2.90  10-10 m2/s Mathematical model Specific energy consumption (SEC) in freeze The measured and predicted moisture ratio drying process (MR) from different models is shown in Figure 7, Table 5 shown comparison of the SEC n and the parameter of the Newton, Page, Modified used in each experiment. In RUN 1, 2 and 3 the Page, Henderson-Pabis, specific energy consumption did not differ, because

64 Journal of Agr. Research & Extension 30(3) (Suppl.): 57-67

1.0 1.0

Experiment 0.8 Experiment 0.8 Page Model Newton Model

0.6 0.6

MR MR MR 0.4 0.4

0.2 0.2

0.0 0.0 0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28 time ,hr time , hr

1.0 1.0

0.8 Experiment 0.8 Experiment Modified Page Model Henderson and Pabis Model

0.6 0.6

MR MR 0.4 0.4

0.2 0.2

0.0 0.0 0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28 time, hr time, hr

1.0 1.0

0.8 Experiment 0.8 Experiment Wang and Sing Model

0.6 0.6

MR MR 0.4 0.4

0.2 0.2

0.0 0.0 0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28 time, hr time, hr

Figure 7 Measured and predicted moisture ratio (MR) from models

65 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Table 5 Coefficients of models obtained in the freeze drying of mangoes

2 Model k n a b c R2 RMSE  Newton 0.121 - - - - 0.958 0.064038 0.00410 Page 0.032 1.591 - - - 0.998 0.016354 0.00026 Modified Page 0.115 1.591 - - - 0.998 0.016215 0.00026 Henderson and Pabis 0.127 - 1.057 - - 0.968 0.060084 0.00361 Logarithmic Model 0.104 - 1.131 - - 0.089 0.979 0.050540 0.00255 Wang and Sing - - -0.087 0.002 - 0.964 0.060364 0.00364

Logarithmic and Wang and Sing models 6.8% and the specific energy consumption was for the drying process and summarized in Table 253.07 kWh/kg. The average drying rate was 2 6. The R values indicates that Page and Modified 0.294gH2O/g dry mass-h and the effective moisture Page models fit reasonably well with the experimental diffusivity coefficient from linear equation method data. Root means square error (RMSE) of Page ranged from 5.54  10-11to 2.90  10-10 m2/s. The and Modified Page model is 0.016354 and 0.016215, Modified Page model was found to be the better respectively. It can be seen that the Modified Page model, with coefficient of determination of 0.998, model performed better than the Page model. It is reduced chi-square value of 0.00026 and RMSE at recommended that the Modified Page model 0.016215. These parameters can be used as guide should be used to predict the freeze drying of to scale up the manufacturing in SME scale mangoes. both in production and investment.

Conclusion Acknowledgements

The optimal conditions were in the secondary The research was supported by the Graduate drying stage at 10oC for 6 hours at the pressure School Maejo University. The authors thank the of 20 Pa. After drying, the quality of mangoes was Faculty of Engineering and Agro-Industry, Maejo determined. the value of hardness was 6.1834 N University for research facilites and to Professor and the water activity was 0.276. The colors L*, a* Siriwat Wongsiri for the helpful suggestions and and b* value were 79.86, 4.29 and 53.62, respectively. critical reviewing of this manuscript. The final moisture content of the product was

66 Journal of Agr. Research & Extension 30(3) (Suppl.): 56-67

Reference Sharma, G.P. and S. Prasad. 2006. Specific energy consumption in microwave drying Crank, J. 1975. Mathematics of Diffusion 2nded. of garlic cloves. J. Energy. 31: 1921-1926. London: Oxford University Press. Sivakumar, D., Y. Jiang and E.M. Yahia. 2011. Dayang, B., F. Ahmad and R. Mohd. 2012. Maintaining mango (Mangifera indica L.) Drying characteristics of the borneo fruit quality during the export chain. canarium Odontophyllum (dabai) fruit. J. Food Research International. J. American J. Agricultural and 44: 1254-1263. Biological Sciences. 7: 347-356. Suchada, C., M. Waraporn, S. Suwaree and Doymaz, I. 2007. The kinetics of forced T. Yupaporn. 2002. Development of convective air drying of pumpkin slices. freezed-dried mango and mangosteen J. Food Eng. 79: 243-238. by using freeze dryer. Proceedings of Nathdanai, H. and C. Sanguansri. 2011. the conference on Science and Influence of collapsed structure on Technology of Thailand. stability of  -carotene in freeze-dried Tunde-Akintunde T.Y., G.O. Ogunlakin. 2011. mangoes. J. Food Research Influence of drying conditions on the International. 44: 3188-3194. effective moisture diffusivity and energy Othman, M.Y., A. Fudholi, K. Sopian, requirements during the drying of M.H. Ruslan and M. Yahya. 2012. pretreated and untreated pumpkin. Drying kinetics analysis of seaweed J. Energ Convers Manage. Gracilaria cangii using solar drying 52(2): 1107-1113. system. J. Sains Malaysiana. Tyree, M.T. 1970. The symplast concept: 41: 245-252. A general theory of symplastic transport Rangel, M., J. Welti, V. Córdova, G. Cerón, according to the thermodynamics of M. Cerón and F. Anguebes. 2011. irreversible processes. Estimation of Moisture Sorption Isotherms J. of Theoretical Biol. 26: 181-214. of Mango Pulp Freeze- dried. Wang, N., Brennan, J.G. 1995. A mathematical J. International Journal of Biology and model of simultaneous heat and moisture Biomedical Engineering. 5: 18-23. transfer during drying of potato. Robles-Sánchez, M., H. Astiazarán-García, J. of Food Engineering. 24: 47-60. O. Martín-Belloso, S. Gorinstein, E. Alvarez-Parrilla, L.A. de la Rosa. 2011. Influence of whole and fresh-cut mangointake on plasma lipids and antioxidant capacity of healthy adults. J. Food Research International. 44: 1386-1391. 67 Journal of Agr. Research & Extension 30(3) (Suppl.): 68-79

Fixed Deep-beds Drying of Black Pepper: A Comparative Study between a Normal Airflow and Reverse Airflow

Phirunrat Thaisamak*, Wipa Teppinta, Chanawat Nitatwichit, Jatupong Varith and Somkiat Jaturonglumlert Division of Food Engineering, Faculty of Engineering and Agro-Industry, Maejo University, Chiang Mai, Thailand 50290 *Corresponding author: [email protected]

Abstract

The purpose of this research study was to compare normal airflow and reversal airflow in fixed deep-bed drying of black pepper. The deep-bed model has temperature and moisture ratios which varied with depth and drying time. The batch drying process was divided into two parts. The first part included data collection of the temperature and moisture distribution, which was affected by two independent variables, the depth (at 20, 30 and 40 cm) and types of airflow (normal airflow vs. reverse airflow every 1 hour). The second part was to fit the experimental data to estimate dimensionless variable to be used in mathematical model to explain drying behavior. It was shown that temperature and moisture ratios of normal airflow and reverse airflow were similar in deep-bed model. The airflow reverse process could be used for drying at any depth of the drying bed, which the physical property was close to the standard and can reduce the drying cost. The model explained the behavior of drying process in nearby boundary condition and can be applied to adjust the dryer and drying process in the industry.

Keywords: fixed deep-bed drying, black pepper, reversal airflow, normal airflow

Introduction and packaging. Reversal airflow drying was developed from batch type drying which the Black pepper (Piper nigrum L.) is popular drying air flow is applied in one direction for some in healthy and spicy herb group. The product of time and then the direction of the flow is reversed pepper gives an annual export value (about 30-60 for the next period that may be repeated several million baht) to country (Office of the Agricultural times until the drying was completed. From this Economics, 2012). Drying is the method to prolong method, the hot air can go through and distribute shelf life for high humidity product as well as to to the product in both directions. reduce the weight of product, cost of instrument

68 Journal of Agr. Research & Extension 30(3) (Suppl.): 68-79

Fixed-bed drying is a complex process of black pepper and study the feasibility of which is used for black pepper in the industry. modeling for the types of airflow (normal airflow The possible option is stirring by experience and reversal airflow) by comparing the moisture workers during drying at regular intervals. This and temperature ratio vary with time and depth. problem incurs high cost and labor intensive. Lastly the optimal depth of bed was determined Thus, the reversal airflow drying was used to by fixed deep-bed model to apply in the industry reduce these problems, the drying time, heat loss scale. Many researches in the field of mathematical between stirring time and cost of labor (Janjai et al., modeling have studied the empirical model which 2011). However the designer process including or could explain the varying behavior of drying with modification of the dryer was of high cost, time time only. consuming and must take longer to find the optimal condition. Mathematical modeling is one of the most Materials and Methods important methods in food drying process which deals with the behavior of drying by moisture and The mathematical model of fixed deep- temperature distribution. This method reduced bed drying was validated for the experimental the number of trial and cost of material. The result data of black peppers by comparing the behavior of mathematical model could be applied to adjust of drying varied with two independent variables drying process and was to modify the dryer. i.e. the depth of bed (20, 30, 40 cm) and the type Fixed deep-bed drying model was developed of airflow (normal airflow and reverse airflow). In to describe the heat and mass transfer during this work, the bulk density of black pepper was drying. This model involves a few variables and about 550 kg/m3. The temperature of drying air parameters for simulation such as the drying was 100๐C. The initial moisture content which constant of thin-layer, thermodynamic properties determined by hot air oven method was 0.17 dry of air and grain. In this research, a fixed deep-bed basis and the average air velocity at the entrance model developed by Barre et al. (1971) was used was 2.25 m/s. The implement of dryer shown in to simulate numerically the fixed deep-bed drying Figure 1.

1. Liquefied Petroleum Gas 1 2. Temperature controller 4 2 3 3. Blower 1/3 hp 5 4. Drying bin 5. Set of measurement

Figure 1 Implement of dryer in drying process

69 Journal of Agr. Research & Extension 30(3) (Suppl.): 69-79

To simulate the model, the drying air temperature, relative humidity and moisture content were determined throughout the experiment. K type thermocouple connecting with the data logger (Agilent technology 34970A) was used to measure the air temperature in the drying bin. Relative humidity of ambient air and drying air were measured with a hygrometer. The moisture content was determined with hot air oven method by Figure 2 (1) Position of moisture content and sampling black pepper from drying bin at designated temperature used to collect simulation position. The experimental data were periodically data; (2) Tray used to contain material; collected at 30 minutes interval until the drying (3) Tray containing black pepper and was completed. The position of thermocouples were position of thermocouple for collecting periodically collected at 30 minutes interval for temperature profile. measurement temperature in the drying bin was defined three positions, and each position was The patterns of airflow in both types are divided into five sub-layers along at the depth of shown in Figure 3. In normal airflow, hot air was bed (at x=1, 2, ..., 5). The position of moisture content applied from the bottom to the top of black pepper was also defined two positions. The position of bed. (open valve A close valve B) For reverse measurement in drying bin is shown in Figure 2 airflow, was applied (hot air) from the bottom to for example T11 was the first position of thermocouple the top with interval time set then was reversed and sub-layer x=1. In the study of reversal airflow, (the hot air) from the top to the bottom (close the direction of airflow was reversed 1-hour at valve A open valve B), the reverse was repeated interval by repeating the reverse direction until the until the process was reached. desired moisture content was reached (about 0.07 on dry basis).

(1) (2) Figure 3 Operation of the fixed-bed dryer using normal airflow (1), reversal airflow (2) 70 Journal of Agr. Research & Extension 30(3) (Suppl.): 68-79

Fixed deep-bed model With Ta0 is inlet temperature of the drying

Fixed deep-bed drying is a non-stationary air; Tam is ambient air wet bulb temperature; is process Moisture content and temperature depended coordinate along bed depth, (m) and H is Height on the depth “ x ” of the bed and on the drying of transfer unit. Assuming the inlet temperature of time “ t ”. Aregba et al. (2006) was proposed drying air is constant (equal to ), as a mathematical model of fixed deep-bed by define for temperature, the moisture content is given by the moisture content of product as M p (x,t) .

The drying air temperature in bed as Ta (x,t). M (0,t)  M Initial p eq  eKt (4) The following assumption was made in deriving M p (x,0)  M eq the equation: The sensible heat required to raise temperature is negligible compared to latent heat for moisture vaporization, The grain bulk density Boundary M p (x,0)  M pi (5) and latent heat vaporization are constant.

The initial moisture content in the bed is equating all over the bin. Hukill (1954) was presented energy With M pi is the initial moisture content of balance equation by the mention assumption. the product, K is thin-layer drying constant, and is time. T M  V Cp    h (1) da a0 da x dp fg t Dimensionless fixed deep-bed model The definition of some dimensionless variable in relation of fixed deep-bed model. In  With da is the density of drying air. relation of initial and boundary conditions was 3  (kg/m ); dp is the bulk density of product, expressed two unknown variables. Aregba et al. [4] 3 Cp (kg/m ); dp is specific heat of drying air,(J/kg); presented the technique involved in this process h V fg is latent heat vaporization of water, (J/kg); a0 corresponding to the theory of transfer units. is inlet velocity of drying air, (m/s) The relation 1 Given H is the height of transfer unit and K is the which has been tentatively proposed is a sample drying constant of thin-layer kinetic equation. form and was improved by using the exponential Equations 1-4 constitute the differential equation form. Hukill further assumed the following initial system of fixed deep-bed. The dimensionless and boundary conditions, for drying temperature: variables are defined as follow:

x T (x,0) T  a am  e H Initial (2) M p (x,t)  M eq Ta (0,t) Tam MR(x,t)  (6) M pi  M eq

Boundary Ta (0,t) Ta0 (3)

71 Journal of Agr. Research & Extension 30(3) (Suppl.): 68-79

T (x,t) T TR(x,t)  a am (7) The fixed deep-bed model was in good agreement T T ao am and acceptable with high R2 and low RMSE value and % error (Narong et al., 2011).   Kt (8) Results and Discussion

x The operation of drying process was (x)  (9) H found to vary with the depth of bed and type of airflow. In this section, in the operation at 40 cm depth of bed is present, with defined Mp is the M With eq is equilibrium moisture content of average moisture content at two positions and Ta black pepper, calculated by Modified Oswin equation is the average drying air temperature at three from the study, that is M eq  (0.150  0.004(T) positions, and the average value was divided by 0.647  RH   0.0000353T 2 )   0.020 . Equations sub-layers. When the experimental result between 1 RH  normal airflow and reverse airflow was compared, 6-9, the depth, x and time, t are expressed in it was found that in normal airflow, the moisture dimensionless term , . The moisture and content in each sub-layer tended to decrease temperature ratios for contain drying time and bed continuously from sub-layer 1 to 5 when was depth can be calculated as follow. applied hot air at the bottom of bed. Although stirring can distribute the hot air, the change of x e e H moisture content varied with the depth of bed. The MR,    e  e 1 x e H  eKt 1 moisture content in each sub-layer was similar when   0,   0 (10) when the drying time passed 150 minutes. In reverse airflow, the moisture content was different in e eKt each sub-layer that is, the moisture content at TR,      x e  e 1 Kt sub-layer 1 was low when dry air was applied at e H  e 1 when   0,   0 (11) the bottom. Similar, sub-layer 5 had low moisture content when hot air was applied reversely at the Non-linear regression with SPSS program top of bed. Moisture content tended to decrease was used for the analytical transfer units. After rapidly when the drying time passed 90 minutes. that, the equations 10, 11 with estimated values The temperature in the normal airflow from analytical transfer unit were used to predict tended to increase continuously from bottom to top of bed and was higher at sub-layers 1, 2 than the moisture and temperature ratio. In this work R2, RMSE and %error were used to validate at other sub-layers. Drying air temperature was between the predicted and experimental results. stable when the drying time passed 150 minutes 72 Journal of Agr. Research & Extension 30(3) (Suppl.): 68-79 and was constant until the drying was completed. this type of airflow, the profile of temperature was In the reversal airflow the moisture content swinging at the beginning of drying and converge decreased as the temperature increased which similar when the drying time passed 120 minutes. was high at sub-layers 1, 2 when hot air was Figure 4 shows the change of moisture content applied at bottom of bed and was high at sub- and temperature profile in the bed in normal and layers 5, 4 when the reversing time took place. In reverse airflows.

0.18 Mp (Normal airflow) at 0.18 Mp (Reverse airflow) at x=1 x=1 0.16 0.16 Mp (Reverse airflow) at x=2 Mp (Normal airflow) at 0.14 x=2 0.14 Mp (Reverse airflow) at x=3 0.12 Mp (Normal airflow) at 0.12 Mp (Reverse airflow) at x=4 0.10 x=3 0.10 Mp (Reverse airflow) at x=5 0.08 0.08

0.06 0.06

Moistue content (d.b) Moistue content (d.b) Moistue content 0.04 0.04 0.02 0.02 0.00 0.00 0 30 60 90 120 150 180 210 0 30 60 90 120 150 180 210 Time (min) Time (min)

100 100

80 80

60 Tp (Normal airflow) at x=1 60 Tp (Reverse airflow) at x=1 Tp (Normal airflow) at x=2 Tp (Reverse airflow) at x=2 40 40

Tp (Normal airflow) at x=3 Tp (Reverse airflow) at x=3

Temperature (celcius) Temperature Temperature (celcius) Temperature Tp (Normal airflow) at x=4 Tp (Reverse airflow) at x=4 20 20 Tp (Normal airflow) at x=5 Tp (Reverse airflow) at x=5 0 0 0 30 60 90 120 150 180 210 0 30 60 90 120 150 180 210 Time (min) Time (min)

(1) Normal airflow (2) Reverse airflow

Figure 4 Moisture and temperature profiles during drying at the condition of 40 cm with average moisture content in three positions and average temperature in two positions Mp = average moisture, Ta = average temperature

73 Journal of Agr. Research & Extension 30(3) (Suppl.): 69-79

From the experiment on drying, the solution analytical result of transfer units divided by moisture of fixed deep-bed model is given by equations 10, and temperature ratio, statistic analysis are shown 11. The result of the dimensionless moisture and in Table 1. temperature ratio fitted the predicted result. The

Table 1 Value of transfer units and statistic analysis

Depth of MR TR drying bed Condition K H R2 RMSE %Error K H R2 RMSE %Error (cm) Normal airflow 0.007 -1.985 0.935 0.0621 8.419 0.004 0.037 0.810 0.1377 7.679 20 Reverse airflow 0.007 1.894 0.889 0.0711 10.260 0.004 0.036 0.851 0.1233 8.630

Normal airflow 0.007 3.275 0.971 0.0841 5.331 0.002 0.074 0.859 0.1785 7.813 30 Reverse airflow 0.006 5.029 0.932 0.0530 7.473 0.002 0.073 0.789 0.1342 12.484

Normal airflow 0.005 3.043 0.977 0.0341 4.620 0.001 0.138 0.831 0.1556 17.959 40 Reverse airflow 0.005 15.026 0.958 0.0069 5.184 0.001 0.100 0.812 0.0936 11.628

The predicted result for variation of moisture temperature ratio. From statistic analysis the best and temperature ratios under different drying fitted with fixed deep-bed model was 40 cm, and conditions were obtained and compared with the the model was in good agreement with moisture experimental result. From Table 1, the minimum ratio and over estimated in some conditions of (R2) between both types of airflow when considering temperature ratio. moisture ratio was 0.971, 0.889 respectively, and Figure 5 shows comparison between the when considering temperature ratio was 0.810, predicted result and experimental results of moisture 0.789. It was shown that the simulate result of ratio between normal airflow and reversal airflow 40 cm was the best fitted (R2=0.977, 0.958). of black pepper with varying depth. (The moisture The minimum value for RMSE were 0.0341, ratio of normal and reversal airflows gave similar 0.0069 respectively when considering moisture result in each condition of drying. The model ratio, and there were 0.1233, 0.0936 when provided good agreement between experimental considering temperature ratio. The maximum value and predicted result and good prediction when the for % Error were 8.14, 10.26 respectively when run duration was less than 60 min. considering moisture ratio and 17.59, 12.63 in

74 Journal of Agr. Research & Extension 30(3) (Suppl.): 69-79

1.00 1.00

0.80 0.80

0.60 0.60

MR MR 0.40 0.40

0.20 Experimental 20 cm (Normal airflow) 0.20 Experimental 20 cm (Reverse airflow) Fixed deep-bed model Fixed deep-bed model 0.00 0.00 0 30 60 90 120 150 0 30 60 90 120 150 Time (min) Time (min) 1.00 1.00

0.80 0.80

0.60 0.60

MR MR 0.40 0.40

0.20 Experimental 30 cm (Normal airflow) 0.20 Experimental 30 cm (Reverse airflow) Fixed deep-bed model Fixed deep-bed model 0.00 0.00 0 30 60 90 120 150 0 30 60 90 120 150 180 Time (min) Time (min) 1.00 1.00

0.80 0.80

0.60 0.60

MR MR 0.40 0.40

0.20 Experimental 40 cm (Normal airflow) 0.20 Experimental 40 cm (Reverse airflow) Fixed deep-bed model Fixed deep-bed model 0.00 0.00 0 30 60 90 120 150 180 210 0 30 60 90 120 150 180 Time (min) Time (min)

Figure 5 Comparison of the moisture ratio of experimental and fixed deep-bed model: normal airflow ๐ (left) and reverse airflow (right) at Tao=100 C , Bed depth = 20, 30, 40 cm

75 Journal of Agr. Research & Extension 30(3) (Suppl.): 68-79

1.00 1.00

0.80 0.80

0.60 0.60

TR TR 0.40 0.40 Experimental 20 cm (Normal airflow) Experimental 20 cm (Reverse airflow) 0.20 Fixed deep-bed model 0.20 Fixed deep-bed model

0.00 0.00 0 30 60 90 120 150 0 30 60 90 120 150 Time (min) Time (min) 1.00 1.00

0.80 0.80

0.60 0.60

TR TR 0.40 0.40

0.20 Experimental 30 cm (Normal airflow) 0.20 Experimental 30 cm (Reverse airflow) Fixed deep-bed model Fixed deep-bed model 0.00 0.00 0 30 60 90 120 150 0 30 60 90 120 150 180 Time (min) Time (min) 1.00 1.00

0.80 0.80

0.60 0.60

TR TR 0.40 0.40

0.20 Experimental 40 cm (Normal airflow) 0.20 Experimental 40 cm (Reverse airflow) Fixed deep-bed model Fixed deep-bed model 0.00 0.00 0 30 60 90 120 150 180 210 0 30 60 90 120 150 180 Time (min) Time (min) Figure 6 Comparison of the temperature ratio of experimental and fixed deep-bed model: normal ๐ airflow (left) and reverse airflow (right) at Tao = 100 C, Bed depth = 20, 30, 40 cm.

76 Journal of Agr. Research & Extension 30(3) (Suppl.): 68-79

Figure 6 shows the comparison between equilibrium iso-therm equation at relative humidity the predicted and experimental result of temperature above 90%, (3) the error in the measurement of ratio between normal airflow and reverse airflow input parameters and actual performance of dryer with varying bed depth. The temperature ratio of (Zare et al., 2006) both types was not in good agreement in some The physical property of final product by conditions. The model provided good prediction in reversal airflow drying was close to the standard normal airflow and some tests of reverse airflow, which the property in general drying industry such but it was somewhat overestimated at the depth as water activity, moisture content and color of was more than 40 cm, although the discrepancies product was used. The final moisture content in between the predicted and experimental value this test was close to the standard of industry, i.e. were acceptable. Therefore, the deep-bed dryer about 0.07. Contamination by microorganisms in model describes the drying process with good the final product was examined and analyzed in accuracy in both types of airflow. The main errors laboratory. It was found to pass and acceptable between the predicted and experimental data are by the standard of drying industry [8]. The result probably due to (1) lack of the estimate transfer of physical property was shown in Table 2. The unit, (2) insufficient precision of the moisture before and final product was shown in Figure 7.

Table 2 Analysis of physical properties

Condition Drying Color Water activity Final Initial Final Bed thickness Airflow time L* a* b* moisture content weight weight (cm) (min) (d.b.) (kg) (kg) Normal airflow 150 40.20 4.65 15.96 0.465 0.066 50.00 45.70 20 Reverse airflow 150 36.81 5.24 14.40 0.403 0.068 50.00 45.70 Normal airflow 150 39.44 4.23 14.98 0.663 0.067 75.00 68.70 30 Reverse airflow 180 36.92 3.84 14.78 0.647 0.068 75.00 68.14 Normal airflow 210 36.88 4.73 13.65 0.426 0.068 100.00 90.60 40 Reverse airflow 195 37.18 4.77 14.66 0.494 0.068 100.00 90.15

The drying behavior of black pepper for be contained at the depth more than 20 cm which bed depth of 20, 30 and 40 cm were similar. The have no effect on the properties of final product. prediction of model can explain the behavior in So the operating cost decreased with increased every bed depth. So, in pragmatic drying, 30-40 bed depth. It is suggested that the dryer with cm of bed height was optimal in the industrial reverse airflow should be operated for economic scale. In each batch of dryer, the product could application. The fixed deep-bed model reduced

77 Journal of Agr. Research & Extension 30(3) (Suppl.): 69-79 cost in the laboratory to ensure optimization. The to adjust the dryer and drying process in the model give good to explanation on the behavior of industry. drying and is useful for the simulation of drying phenomena and dryer performance. This type of Acknowledgements model can be used by decision support systems for the investigation on drier design problem. We would like to thank the research project “Operation Cost Reduction for Industrial Pepper Power Drying with Alternating Hot-Air During Drying Process” funded by Wipa Tappinta, Thailand. We would also like to thank the executives and employees of NithiFoods Co., Ltd which provide facilities for this work.

Before After Reference

Figure 7 Black pepper in drying process: Aregba, A.W., P. Sabastian and J.P. Nadeau. before drying and after drying 2006. Stationary deep-bed drying: A comparative study between a Conclusion logarithmic model and a non-equilibrium model. J. of Food Engineering. A comparison between a normal airflow 77: 27-40. and reverse airflow was investigated by fixed Barre, H.J., G.R. Baughaman and M.Y. Hamdy deed-bed model which was derived from Hukill’ s 1971. Application of the logarithmic model. The variables of the model were formulated model to crossflow deep-bed drying. and estimated with non-linear regression. It was Transection of the ASAE. found that the deep-bed model gave similar result 14(6): 1061-1064. which could explain temperature and moisture Hukill, W.V. 1954. Grain drying. In J.A., ratio in both types of airflow. The reverse airflow Anderson, A.W. Alcock (Eds.) Storage of could be used at all depth of drying bed, but was Cereal Grain and Their Products. St. underestimated in the bed depth more than 40 Paul, Minn: American Associate Cereal cm. The simulation model could be used as a tool Chemical. to optimize drying process and decreased cost Narong, U., B. Panupong and H. Wanphen. 2011. of optimization in drying process. The model was Models of pepper drying using hot air useful for drier design to predict and explain oven. J. Agricultural Sci. behavior of drying process which can be applied 42(3): 533-536.

78 Journal of Agr. Research & Extension 30(3) (Suppl.): 69-79

Janjai, S., N. Lamlert, B. Mahayothee, Thai Industrial Standards Institute, Thailand. P. Sruamsiri, M. Precoppe, B.K. Bala 2004. TCPS 491-2547 (2004) (Thai): and Muller J. 2011. Experimental Ground black and white pepper us and simulation performances of a batch– government documents. [Online]. type longan dryer with air flow reversal Available https://law.resource.org using biomass burner as a heat source. (2013 March 10). J. Drying technology. 29: 1439-1451. Zare, D., S. Minaei, M. Mohammadzadeh. Office of the Agricultural Economics. 2012. and M. Khoshtagaza. 2006. Computer Import-export statistics, Ministry of simulation of rough rice drying in a batch Agricultural and Cooperative. dryer. International Journal of Energy [Online]. Available http://www.oae.go.th Conversion and Management. (2013 February 20). 47: 3241-54.

79 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87

Operation Cost Reduction for Industrial Pepper Powder Drying with Alternating Hot-air during Drying Process

Wipa Teppinta, Jatuphong Varith*, Somkiat Jaturonglumlert, Phirunrat Thaisamak and Chanawat Nitatwichit Program in Food Engineering, Faculty of Engineering and Agro-industry, Maejo University, Chiang Mai, Thailand 50290 *Corresponding author: [email protected]

Abstract

Drying is an important process to control microorganism in black pepper powdering process. In industry, the traditional hot-air (THA) drying on pepper is accomplished using one-direction air flow which requires thorough stirring labor to uniform temperature of pepper. To minimize labor cost, the alternating hot-air (AHA) drying which air-flow can be forced to upward or downward direction is proposed. The objectives of this research were to study the AHA drying process for black pepper and to compare the operation cost between alternating and THA drying process. The experiment was conducted to dry black pepper of 50-100 kg/batch from 16-18% to 7% (dry basis) using 100°C hot air in a hot-air drying cabinet which can alternate the hot-air flow onto upward or downward direction. Thickness of pepper bed (20, 30 and 40 cm) and drying methods (hourly for AHA drying, and ½ hourly laboring stir for THA drying) were two important factors in this study. Results show that different drying methods using THA and AHA yielded similar results in term of drying characteristics and drying time. Statistical analyses indicate that drying curve and drying rate between both methods were not significantly different (P>0.05). However, AHA drying employed less labor man-hour than THA drying by 31%, reflecting less operation cost per batch, due to no need for stirring process to uniform temperature of the black pepper during drying process.

Keywords: Black pepper, alternating hot-air, operation cost

Introduction and for export by 92% (Department of Agricultural Extension, 2009). The consumption of pepper can Pepper (Piper Nigram L.) is an important be in form of fresh pepper and powdered pepper. crop which is used as a basic cooking spice and Drying is one of the oldest and most food ingredient worldwide. Pepper crop is grown important food preservation methods by applying in intensively in North Eastern area of Thailand heat and mass transfer principles to remove moisture which is used for domestic consumption by 73% from the drying food materials (Achariyaviriya et al.,

80 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87

2007). In the industry, hot-air drying is the important drying industry. Varith et al. (2009) reported on step to reduce moisture and lower water activity to AHA drying of longan fruit that the AHA drying the safety limit. The fixed-bed typed hot-air drying provided not only labor cost reduction, but also is among the most popular and traditional hot-air less operating cost and less fruit damage for the final (THA) drying methods for food industry which is product (from 10% to 2%). Thus, the objectives of continuously used for pepper powdering industry. this research were to study an alternating hot-air For industrial THA drying, the pepper seed is drying process for black pepper drying and to placed inside the drying cabinet with capacity compare the operation cost between AHA and ranging from 500 to 3,000 kg. The hot-air stream THA drying processes. is forced upward from bottom to top of the bed through pepper seed and then released the moisten Materials and Methods hot air out from the cabinet. Even though it has an advantage of simplicity, the THA drying process Dried black pepper seed used in this requires the massive laboring for stirring the pepper research was provided by Nithi Foods Co. Ltd., to distribute and uniform heat throughout the bed for Thailand. Physical properties of black pepper were: which it becomes the major cost for pepper powder bulk density of 550 kg/m3 and average moisture processing. content of 16.5% (dry basis). The experiment was The alternating hot-air (AHA) drying is a conducted to dry the black pepper seed of 50-100 modified version of fixed-bed type hot-air cabinet kg/batch to 7% (dry basis) using 100ᵒC hot air. where the hot air stream is forced either upward The AHA dryer and the data collecting positions were or downward through the drying material with the schematically illustrated in Figure 1. Thickness of airflow alternating mechanism. The AHA drying black pepper bed (20, 30 and 40 cm) and drying gives advantage over the THA drying on less methods (hourly alternating air flow and traditional labor to deal with drying process, which is of high hot-air with ½ hourly laboring stir) were two important potential to replace the stirring process in pepper factors in this study. All experiments were triplicated.

Figure 1 Schematic diagram of AHA dryer and the positions of data collection. The ∆ represents temperature and moisture content data collecting position. 81 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87

Principle of the AHA dryer shown in Page Model MR=exp(-ktn) (1) Figure 1(a) was based on alternation of the air flow either by upward or downward direction. Modified Page ModelMR=exp[-(kt)n] (2) During drying, fresh air flows through a LPG nozzle-typed heater to generate hot-air stream. For quality of dried pepper, water activity The hot air was then forced through black pepper measured by AquaLab model TE3 (Decagon bed either from bed bottom (2) to bed top (3), or Devices, Inc., USA), color indices L*a*b* measured vice versa. The exhausted hot-air was then by Spectrophotometer model Miniscan XE (Hunter released from the dryer on bed top (4) or bed Labs, Inc., USA) and specific energy consumption bottom (5) with the directional switching valve. (excluding energy loss of drying cabinet) were analyzed as the responses of the drying process. Mathematical modeling of drying curves The operation cost was estimated from the labor In the experiment, temperature and man-hour, wage of 37.50 Bahts/h and loading moisture content were collected every 30 min capacities of black pepper. using with thermocouples type K connected to data acquisition system model 34970A (Agilent Results and Discussion Technology, Inc., USA). Drying characteristics was then analyzed using Page’s and Modified Page’s Effects of black pepper bed thickness for model for drying constants and drying rates alternating air flow according to Equation 1 and Equation 2. Regression Effects of three different thicknesses of correlation coefficient (r2), error, RMSE and black pepper bed (20, 30 and 40 cm) on the analysis was performed with analyses of the %error drying characteristics and drying rate curves with as primary criteria for best fitted model. AHA drying were shown in Figure 2a and 2b,

0.20 0.40

20 cm 20 cm

dry dry /g

.h) 30 cm 0.15 30 cm 0.30

40 cm 40 cm

water

g

dry solid dry /g

0.10) 0.20

water

solid g 0.05 0.10

Moisture content ( content Moisture 0.00 0.00 Drying rate Drying rate ( 0 1 2 3 4 0.00 0.05 0.10 0.15 Time (hr) Moisture content (g /g ) water dry solid a. b. Figure 2 Effects of bed thickness on the drying curve (a.) and drying rate (b.) for black pepper with AHA drying

82 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87 respectively. It was found that the moisture range of 0.28 To 0.02. The results were found in content decreased as drying process was carried similar trend with the study of other fruit drying on with different bed thickness. Moisture content such as longan (Varith et al., 2009; Janjai et al., of black pepper decreased from 16% to 7% dry 2011). basis (0.16 to 0.07 gwater/gdry solid) within 1.5, 3 and 3.25 hrs for thickness of black pepper bed of 20, Effects of drying methods on drying characteristics 30 and 40 cm, respectively. The thicker bed required Comparison between THA and AHA drying, more moisture removal, resulting in longer drying drying curves and temperature at mid-point of time. Drying rate of AHA pepper drying fell in the drying bed were analyzed as shown in Figure 3.

0.20 120 0.20 120

100 100 0.15 0.15 80 80

0.10 60 0.10 60 Temperature Temperature (ºC)

40 40 Temperature (ºC) 0.05 0.05 Moisture Moisture content (gwater/gdry solid) Tf-M Af-M 20 Tf-M Af-M 20

a 20 Moisture content (gwater/gdry solid) Tf-T Af-T Tf-T Af-T b 30 cm 0.00 0 0.00 0 0 1 2 3 0 1 2 3 4 Time (hr) Time (hr) 0.20 120

100 0.15 80

0.10 60

40 Temperature (ºC) 0.05 20 Moisture Moisture content solid) (gwater/gdry Tf-M c 40 cm 0.00 0 0 1 2 3 4 Time (hr)

Figure 3 Effect drying methods on drying characteristic with 3 different bed thicknesses: (a) 20, (b) 30 and (c) 40 cm. Abbreviations Tf and Af represents THA and AHA dryings, respectively. Notations M and T represent moisture content and temperature of pepper bed.

83 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87

0.400 TF-20 cm

.h) AF-20 cm 0.300 TF-30 cm

dry solid dry AF-30 cm /g

0.200 TF-40 cm water AF-40 cm 0.100

Drying rate (g rate Drying 0.000 0.00 0.05 0.10 0.15 Moisture content (g / g ) water dry solid

Figure 4 Effect of drying methods on drying rate: Abbreviations TF and AF represents THA and AHA dryings, respectively

Table 1 Prediction model for AHA drying process for black pepper.

Thickness Model k n R  RMSE %Error of bed (cm) Page 0.0534 0.563 0.99 2.32E-05 1.17E-02 1.81 20 Modified Page 0.0055 0.563 0.99 2.28E-05 1.17E-02 1.81 Page 0.0123 0.837 0.99 1.10E-04 2.77E-02 4.27 30 Modified Page 0.0052 0.837 0.99 1.10E-04 2.77E-02 4.33 Page 0.0062 0.951 0.99 1.10E-04 2.96E-02 3.85 40 Modified Page 0.0049 0.951 0.99 1.10E-04 2.96E-02 3.86

The effects of drying methods on moisture pepper bed were similar between THA and AHA. reduction of black pepper were not significanly The drying rates shown in Figure 4 between THA different (P≥0.05). The temperature distribution of and AHA were not significantly different (P≥0.05), pepper bed between two methods slightly which was in a range of 0.02 to 0.30 gwater/gdry solid.h. different, however, the temperature gradient and Both drying rate decreased continuously as moisture fluctuation across the bed with thickness of 40 cm content decreased (Figure 3c) with AHA was noticable. This fluctuation Table 1 shows the results of the statistical was caused by the air flow direction alternation analyses undertaken by Page and Modified Page where the air flow was reversed from upward to models for AHA pepper drying based on R,  downward direction for every 1 hr. However, toward and RMSE. Both models can predict the moisture the end of drying process, the temperature of content to fit with the experimental data with R of 84 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87

0.99. The parameter n was 0.563 for both models characterize the drying process for black pepper while the parameter k was slightly different between with our drying configuration. both models. Thus both models can be used to

Table 2 SEC and quality of dried black pepper with two drying methods.

Thickness of bed Drying method Water activity, SEC Color (cm) aw (MJ/kgwater) L* a* b* 20 THA 0.465 76.71 40.20 4.65 15.96 AHA 0.403 91.82 36.81 5.24 14.40 30 THA 0.663 54.26 39.44 4.23 14.98 AHA 0.647 54.28 36.92 3.84 14.78 40 THA 0.426 49.71 36.88 4.73 13.65 AHA 0.494 54.57 37.18 4.77 14.66

SEC and quality dried black pepper and 30 cm, the SEC was similar. Thus, for our The SEC and quality of dried black dryer, the recommended bed thickness to dried pepper is shown in Table 2. SEC for AHA drying black pepper should be in maximum of 30 cm. (20, 30 and 40 cm) were 91.82, 54.28 and 54.57 For quality of dried black pepper, the

MJ/kgwater while those of THA drying were 76.71, average water activity (aw) of dried whole black 54.26, 49.71 MJ/kgwater, respectively. With bed pepper for the drying test of 20, 30 and 40 cm thickness of 40 cm, the SEC of AHA was slightly was 0.516. The values were in acceptable range higher than that of THA. Since the stirring process of the aw from industry standard (aw=0.600). For in THA drying process allows hot air to flow both THA and AHA, the color of dried black through the bed easier than that of AHA drying, pepper were not significant (P>0.05), thus the overall this results in less power consumption to dry the quality of dried black pepper were acceptable to the pepper. However, with the bed thickness of 20 industry standard.

85 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87

Comparison of operation cost due to much less labor man-hour on stirring Table 3 exhibits the operation and its cost process. The THA drying requires at least 2 for black pepper THA and AHA dryings with bed workers for stirring with total stirring time for 1.67- thickness of 20, 30 and 40 cm. The operation 2.33 hours. On the other hands, the AHA drying cost for AHA drying was 2.68, 2.05 and 1.63 requires only 1 worker to switch the reversal valve Bahts/kg, respectively, which was less than that of with the operation time only 0.1 hour. The total THA drying by 31%, 21% and 31%, respectively. cost increased with thicker bed drying bed as well The major cost saving on AHA over THA drying is (Janjai et al., 2011).

Table 3 Comparison of operation cost between AHA and THA dryings for black pepper

Labor man-hour Operation (hr) 20 cm 30 cm 40 cm THA AHA THA AHA THA AHA Loading 0.67 0.67 0.67 0.67 0.67 0.67 Drying 2.50 2.50 2.50 3.00 3.50 3.25 Air alternation - 0.07 - 0.10 - 0.10 Stirring 1.67 - 1.67 - 2.33 - Unloading 0.33 0.33 0.33 0.33 3.33 0.33 Total 5.17 3.57 5.17 4.10 9.83 4.35

Wage rate (Baht/h) 37.50 37.50 37.50 37.50 37.50 37.50 Cost per day (Baht) 193.75 133.75 193.75 153.75 368.75 163.13 Cost per kg (Baht) 3.88 2.68 2.58 2.05 2.56 1.63

Conclusion Acknowledgements

The AHA drying yields the similar effects on The authors wish to thank The Thailand moisture content reduction, drying rate, dried pepper Research Fund (TRF) for financial support under quality, and SEC compared to the THA. However, grant# MRG545E054. Our gratitude is extended AHA drying provides more benefit in term of to Nithi Foods Co., Ltd. in providing material, operation cost, which is much less than THA by 31%. manpower and financial supports throughout this Information from this study can be further analyzed research. for economic aspect for upscaling the AHA process in industrial pepper powder drying process.

86 Journal of Agr. Research & Extension 30(3) (Suppl.): 80-87

References

Achariyaviriya, A, K. Maneeboon and Janjai S., N. Lamlert, B. Mahayothee, P. Sruamsiri, W. Jeorentong. 2007. Performance M. Precoppe, B.K. Bala and J. Muller. of a Batch Type Dryer for Longan Fruits 2011. Experimental and simulated Drying. pp.381-388. In The Conference performances of a batch-type longan on Mechanical Engineering Network dryer with air flow reversal using biomass of Thailand, 21st. burner as a heat source. Department of Agricultural Extension Thailand. J. Drying Technology. 29: 1439-1451. 2009. Pepper. [Online]. Available Varith, J., R. Chuaviroj, A. Sipitakiat and http://www.doae.go.th (2012 February 13). J. Pingparawalee. 2009. Development of Embeded System for Controlling the Hot air Switching-Typed Longan Dryer. Chaing Mai: Faculty of Engineering and Agro-industry, Maejo University, Chiang Mai.

87 Guide for Authors

Manuscripts submitted for publication should be of high academic merit and are accepted on condition that they are contributed solely to the Journal of Agricultural Research and Extension. Manuscripts, parts of which have been previously published in conference proceedings, may be accepted if they contain additional material not previously published and not currently under consideration for publication elsewhere. Submission of a multi-authored manuscript implies the consent of all the participating authors. All manuscripts considered for publication will be peer-reviewed by independent referees.

Submission checklist Manuscript submission must include title page, abstract, key words, text, tables, figures, acknowledgments, reference list and appendices (if necessary). The title page of this file should include the title of the article, full names, official name and affiliations of all authors, E-mail address, telephone and fax numbers and full postal address of the corresponding author.

Preparation and submission of manuscripts Authors submitting manuscripts for consideration for publication should follow the following guidelines. 1. Manuscript texts must be written using high-quality language. For non-native English language authors, the article should be proof-read by a language specialist before it is sent to Journal. 2. Manuscript texts should not exceed than 10 pages and the combined number of figures and tables. The inclusion of more figures and tables will reduce the word allowance, and vice versa. 3. The manuscript text and tables should be created using Microsoft Word. 4. Manuscript texts should be prepared single column, with sufficient margins (1.0 inch) for editorial and proof-reader’s marks. 12 pt Times New Roman font should be used throughout and all pages numbered consecutively. 5. Abstracts should not exceed than 200 words. About 5 keywords should also be provided. 6. All measures in the text should be reported in abbreviation 7. Tables and figures should each be numbered consecutively. 8. Acknowledgments should be as brief as possible, in a separate section before the references, not in the text or as footnotes. 9. Citations of published literature in the text should be given in the form of author and year in parentheses; (Hoffmann et al., 2001), or, if the name forms part of a sentence, it should be followed by the year in parenthesis; Hoffmann et al. (2001). All references mentioned in the reference list must be cited in the text, and vice versa. The references section at the end of the manuscript should list all and only the references cited in the text in alphabetical order of the first author’s surname. The following are examples of reference writing.

Reference to a journal article: Chowdhury, M.A.H., R. Begum, M.R. Kabit and H.M. Zakir. 2002. Plant and animal residue decomposition and transformation of S and P in soil. Pak. J. Bio. Sci. 5(7): 736-739.

Reference to article or abstract in a conference proceedings: Friedrich, R. and T. Marheineke. 1994. Life cycle analysis of electricity system: methods and results. Proceedings of an IAEA advisory group meeting/workshop Oct. 4-7, 1994. China.

Reference to a book: Rajeshwar, K. and J.G. Ibanez. 1997. Environmental electrochemistry. San Diego: Academic Press. 327 p.

Reference to an edited book: Hill, S.E. 1996. Emultions. pp. 153-185. In Hall, G.M. (ed.) Methods of testing protein functionality. London: Chapman & Hall. Reference to an electronic data source (used only when unavoidable): Supplier/Database name (Database identifier or number)/Item or accession number [Access date] should be included National Economic and Social Development Board (NESDB). 2001. Input-output tables of Thailand. [Online]. Available http://www.nesdb.go.th [8 August 2001].

10. Submission of manuscript includes 3 copies in which one of them must conform to the format of the Journal of Agricultural Research and Extension. The other must be submitted without the author’s name and office with diskette/CD and cover letter to the editor. All should be directed to the editor at the address given. JOURNAL OF AGRICULTURAL RESEARCH AND EXTENSION Office of Agricultural Research & Extension Maejo University

Vol. 30 No.3 (Suppl.) June – September 2013 ISSN 0125-8850

Killng Varoa Mite by Grooming Behavior of Russian and Thai Honey Bees Boonmee Kavinseksan 1-13

Control of Off-flavor Cyanobacteria in Ponds using Nile Tilapia (Oreochromis niloticus) and Charcoal Bioreactor System Redel Gutierrez, Niwooti Whangchai, Khomsan Ruangrit and Tomoaki Itayama 14-28

Medicinally Potential Plant of Anisomeles malabarica (L.) R. Br. in Comparison with a Porometer Rameshprabu Ramaraj and Yuwalee Unpaprom 29-39

Isolation and Identification of Cyanobacteria from a Freshwater Aquaculture Pond in Northern Thailand Dong Xia, Norio Iwami, Korntip Kannika, Chayarat Pleumsumran Sirapran Fakrajang, Chayaporn Teecharernwong, Redel Gutierrez Zhong Junsheng, Niwooti Whangchai and Tomoaki Itayama 40-48

Carbon Footprint of Central Canteen of Mahidol University Salaya Campus, Thailand Sayam Aroonsrimorakot, Chumporn Yuwaree, Chumlong Arunlertaree Rungjarus Hutajareorn and Tarinee Buadit 49-55

Mathematical Model of Freeze Drying on Mango Sakawduan Keawdam, Chanawat Nitatwichit, Jatupong Varith and Somkiat Jaturonglumlert 56-67

Fixed Deep Beds Drying of Black Pepper: A Comparative Study between a Normal Airflow and Reverse Airflow Phirunrat Thaisamak, Wipa Teppinta, Chanawat Nitatwichit Jatupong Varith and Somkiat Jaturonglumlert 68-79

Operation Cost Reduction for Industrial Pepper Powder Drying with Alternative Hot-air during Drying Process Wipa Teppinta, Jatupong Varith, Somkiat Jaturonglumlert Phirunrat Thaisamakand and Chanawat Nitatwichit 80-87