Food Sci. Technol. Res., 19 (3), 479–483, 2013 Note

Sensory Analysis of the Taste of Ayu Sampled from Different River Environments

1* 2 1 1 Noriko Hattori , Asako Uchida , Mitsuaki Sano and Tetuo Murakami

1 Graduate School of Human Life Science, Nagoya Women’s University, 3-40 Shioji-cho, Mizuho-ku, Nagoya 467-8610, 2 Toyota Yahagi River Institute, 2-19 Nishi-machi, Toyota 471-0025, Japan

Received December 18, 2012; Accepted January 22, 2013

It is widely believed that attached diatoms are the best food source for freshwater ayu Plecoglossus altivelis (Temminck et Schlegel) in Japan. However, sensory testing by a panel consisting of 40 untrained tasters and analyses of gut content showed that there was no significant correlation between preference scores for fish taste and the dominance of diatoms in the algal assemblages fed upon by ayu in the Yahagi River, , Central Japan. Ayu that fed on the filamentous blue-green algae Homeothrix varians was evaluated higher for aroma and fat content than those that fed on diatoms. Ayu feeding on diatoms were collected in the lower reaches of the Yahagi River which receive domestic sewage. In the ab- sence of more detailed analyses, exposure to sewage is considered to confer a characteristic aroma on Ayu.

Keywords: attached algae, Ayu, gut content, sensory testing, Yahagi River

Introduction also doubtful. For example, Hodoki et al. (2003) showed that Ayu, Plecoglossus altivelis (Temminck et Schlegel), a there was no significant difference in body length and weight member of Plecoglossidae, is one of the most popular fresh- between two ayu populations that fed on either diatoms or water fishes in Japan. Consumers are particularly fond of its blue-green algae. cucumber- or watermelon-like scent. The fish is typically The relationship between market value of ayu and its grilled with only salt, and is typically eaten whole, including food source also remain unresolved. It is possible that the the gut (Hitomi, 1697). It is believed that the aroma, taste, river environment has a marked impact on the taste of the and size of herbivorous ayu are determined by the food in the ayu as it is the river that directly affects the periphytic algae gut, which typically consists of various species of periphytic and the ayu that live there. In this study, we present the re- algae. Fishermen claim that, compared to green (Chloro- sults of sensory evaluations of the aroma and taste of Ayu, phyceae) or blue-green algae (Cyanophyceae), diatoms and discuss the taster-preference in relation to algal species (Diatomaceae) are the best food for ayu. Fishermen object to composition in the fish gut and the river environment. water pollution and channel modification, because these have a negative effect on the biomass and species composition of Materials and Methods the periphytic algal communities, and subsequently, on the Description of study area We selected the Yahagi River aroma, taste, and size of ayu. in central Japan as the study site because the ayu and the Nevertheless, the direct relationship between fish aroma river environment of the area have been extensively investi- and the food source of ayu has not yet been clarified. The gated by the Toyota Yahagi River Institute since 1994 (Niimi, aromatic components of the body surface of ayu are not 1997). considered to be derived from algae (Hirano et al., 1992), as The Yahagi River rises in Mt. Ookawairi-Yama (1,908 the carnivorous trout Coregonus clupeaformis has a similar m) and discharges into . The main stream is 118 aroma as ayu (Josephson et al., 1983). km-long and drains a catchment area of 1,830 km2. Seven The belief that ayu size is strictly determined by diet is dams exist along the main river and approximately 1,100,000 people in the catchment (Fig. 1). Organic matter originating *To whom correspondence should be addressed. from domestic waste and plankton in water released from the E-mail: [email protected] dams upstream are the main pollution loads. 480 N. Hattori et al.

used for the sensory tests were standardized by selecting only those individuals with body length of 19.9 ± 1.4 cm in. Within 3 h after capture, the captured fish were transferred from river water and individually sealed in plastic bags to avoid contamination of the aroma. Fish were then preserved at −20℃. Sensory testing Samples were kept frozen for approxi- mately two weeks, when sensory tests were performed on 16 September 2006. The 40 untrained panel members consisted of 30 men and 10 women, ranging in age from twenties to seventies (mode: 55). The fish samples were sprinkled with salt and roasted by a chef over charcoal fire, before being cooled to room temperature before testing. One fish from each of the five sampling points was served to each member. We did not determine the sex of the fish that were served, neither did we prescribe which part of the fish should be tasted. After tasting one sample from each of the five sample sites, panel members were asked to score the samples by 10 Km aroma, taste, and fat content, with a score of 1 indicating the top score and a score of 5 indicating the lowest. If they are unable to score the taste, then they were allowed to leave the answer blank. Scores were subjective, which meant that one panel member may favor cucumber-like smell of ayu while another may dislike it. Evaluation of fat content was also subjective. Although concentrations of lipids and fatty acids in ayu have been shown to fluctuate seasonally (Hirano and Suyama, 1983), this variation in fat content has not been examined within the context of the effect on taste. In order to avoid carryover samples, panel members were required Fig. 1. Geographic location of study area in central Japan. to rinse their mouths with mineral water after testing each Trapezoids across the river show dam sites. Yahagi Dam in the up- sample. After three sensory evaluations, we asked them to permost river reaches is the largest dam. Impounded water from the score their general preferences, not as the sum of each sen- dam is released immediately upstream from Sta. Y5. Y1: Aoi-Oohashi Bridge, Y2: Toyota-Oohashi Bridge, Y3: Fusso, sory score, but as an evaluation of the whole sample. Y4: Nishi-Hirose, Y5: Kawaguchi (Toyota City, Aichi Prefecture) Statistical significance was determined by the Newell & MacFarlane and Krammer methods. If the total scores were The Toyota Yahagigawa riv. Fishery Cooperation Union not significant according to the Krammer table, then the taste stocks the river with juvenile ayu using inland and marine was considered preferable. If the Krammer table scores were stocks ; upstream migration of juvenile ayu from the estuary high, then the taste was considered unfavorable. Differences is reported every year during early summer (Takahashi and in preference were elucidated by the Newell & MacFarlane Niimi, 1999). The migrated/stocked ratio of captured adult is method. Valid answers for the general preference, aroma, unknown, but it is assumed to be low. taste and fat content were obtained from 39, 18, 19 and16 of Fish samples Ayu samples were obtained for the senso- the respondents, respectively. ry tests using live decoy over several days in early September Reference of water quality information Water qualities 2006 at five stations along the Yahagi River (Fig. 1b). Since for the Yahagi River reported by Shiragane (2002) were used body mass may affect evaluations of aroma and fat content, in our study to evaluate the river condition. we measured the sizes of some of the fish samples that were Examination of gut contents Five ayu collected at each not used for the sensory tests. A significant correlation was station were subjected to gut content analysis. Gut contents observed between fish length and weight irrespective of of the front digestive tract were emptied on a slide glass. the sampling points (R2 = 0.95, n = 23), indicating that the Since dominant algal species is the filamentous blue-green fish were neither overfed nor lean (pin-head). The samples algae (Homeothrix varians) and it is difficult to count cell Sensory Analysis of the Taste of Ayu Sampled from Different River Environments 481 number, we indicated its relative abundance as a proportion Newell & MacFarlance’s method detected significant dif- of the microscopic field occupied by the alga. This was done ferences for aroma and general preference between Stas. Y1 by diving the microscopic field into a grid of 200 squares (10 and Y4, and between Stas. Y4 and Y5. × 20). The taxonomic category of the algae in each square The statistical analyses showed that fish from Sta. Y4 was then noted. A total of 4,000 grids were examined at each were more preferred than fish from the other stations, espe- sampling station. Nomenclature followed Kramer & Lange- cially Stas. Y1 and Y5. The tests also showed that general Bertalot (1991) for the Bacillariophyceae and Hirose and preference was largely dependent on aroma. Yamagishi (1977) for the other taxa. Digestive tract content Algal species compositions in the guts of ayu are shown in Table 2. Dominant algal species Results were Homeothrix varians (Cyanophyceae) and some species The order of preference Average scores for aroma, of diatoms (Bacillariophyceae) such as Melosira varians and taste, fat content, and general preference are shown in Table Fragilaria spp. The relative abundance of the algae in the gut 1. The mean and standard deviations of general preference (i.e. Cyanophyceae/ Bacillariophyceae) at Stas. 1 − 5 were scores of the fish from stations 1 − 5 were 3.6 ± 1.2, 3.1 ± 1.4, 0.48, 2.81, 3.98, 6.81 and 3.91, respectively. Filamentous 2.6 ± 1.5, 2.3 ± 1.3 and 3.4 ± 1.3, respectively. A significant green algae, such as Cladophora sp., were relatively rare and difference was detected between Stas. Y1 and Y4 by Kram- digestion of these cells was not observed. mer’s method, indicating that the ayu from Sta. Y4 were sig- Water quality in the Lower Yahagi River In contrast to nificantly more preferred over those of the other stations (p < the stations upstream, the water quality at the lowest station, 0.05). Similarly, the fish from Sta. Y1 were evaluated signifi- Y1, was reported by Shiragane (2002) as being polluted by cantly lower than the fish from the other stations (p < 0.05). domestic sewage. Though indices of organic pollution such Compared to fish from the other stations, the aroma and fat as Chemical Oxygen Demand (COD) and Biochemical Oxy- content of fish from Sta. Y4 were also preferred p( < 0.05), gen Demand (BOD) were approximately 2 and 1 mg/L, re- although the response rate for aroma and fat content was not spectively, the mean annual concentration of NH4-N was 0.03 high; 14 out of 18 responses for aroma evaluated Y4 higher mg/L, with levels greater than 0.08 mg/L recorded in winter. than Y1, and 14 out of 16 responses for fat content evaluated Conversely, NH4-N concentrations near Sta. Y4 were less Y4 higher than Y1. No significant differences between sta- than 0.01 mg/L. The effects of the two dams between Stas. tions were observed for taste. Y1 and Y4 are not described in the report.

Table 1. Sensory analyses of ayu captured at different points along the Yahagi River.

General preference Aroma Fat Taste Sta. Y 1 3.6 ± 1.2* 3.7 ± 1.0* 3.6 ± 1.2* 3.4 ± 1.4 Y 2 3.1 ± 1.4 3.1 ± 1.2 2.9 ± 1.5 2.8 ± 1.3 Y 3 2.6 ± 1.5 2.6 ± 1.2 3.1 ± 1.4 3.1 ± 1.6 Y 4 2.3 ± 1.3* 2.0 ± 1.4* 2.1 ± 1.3* 2.2 ± 1.3 Y 5 3.4 ± 1.3 3.7 ± 1.4 3.3 ± 1.1 3.5 ± 1.2

See Figure 1 for the location of sampling stations Y1 to Y5. Results are shown as mean scores with standard deviations. * indicate significant differences (p < 0.05) detected by the Krammer’s method.

Table 2. Algal species composition in the guts of fish caught at five sample stations.

Bacillariophyceae (B) Cyanophyceae (C) C/B Sta. Y 1 4.1 ± 1.7 2.0 ± 1.0 0.48* Y 2 3.6 ± 1.1 10.1 ± 6.7 2.81* Y 3 5.5 ± 2.1 21.9 ± 15.5 3.98* Y 4 4.3 ± 4.9 29.3 ± 11.2 6.81* Y 5 5.4 ± 4.6 21.1 ± 16.6 3.91*

See Figure 1 for the location of sampling stations Y1 to Y5. Results are mean scores for the proportion (%) of squares occupied by algae in an ocular grid relative to the number of squares that did not contain any algae (± standard deviation; n = 4,000 grid samples). * indicate significant differences (p < 0.05) in relative proportions of Bacillariophyceae and Cyanophyceae. 482 N. Hattori et al.

Discussion The sensory testing revealed that the most preferred ayu Acknowledgements We thank members of Yahagi-gawa Tennen fed on blue-green algae (Sta. Y4), and the least preferred fish Ayu Chosa-kai (Research Group for Wild Ayu in the Yahagi River) fed on diatoms (Sta. Y1). Although significant differences for collecting samples for sensory testing. We are also grateful to were also detected in aroma and fat content, these evalua- Dr. Toshiya Yamamoto and Akiko Shiragane of the Toyota Yahagi tions require further analysis as the response rates for taste, River Institute for their helpful advice regarding the river environ- aroma, and fat content were lower than they were for the ment and catching ayu. general evaluation. In addition, we were unable to account for possible differences in taste between male and female References fish, at least for fat content (Joeng et al., 2000). These find- Abe, S., Katano, O., Nagumo, T., Tanaka, J. (2000). Grazing effects ings do not support the widely held belief that a diet of dia- of ayu, Plecoglossus altivelis, on the species composition of ben- toms improves the taste of ayu flesh. Although ayu fed on thic algal communities in the . Diatom, 16, 37-43. blue-green algae such as Anabaena spp. have been reported Abe, S., Iguchi, K., Matsubara, N., Yodo, T., Tanaka, J., Nagumo, to have a grassy smell (Oohashi and Hatama, 2005), Homeo- T. (2003). Benthic algal flora and the algal composition of the gut thrix varians is considered to be a good food source for ayu contents of ayu Plecoglossus altivelis in the estuary of the Choshi in the Yahagi River (Uchida, 2002) and other Japanese rivers River, Mie Prefecture, Japan. Jpn J Phycol, 51, 117-121. (in Japa- (Abe et al., 2003, 2005). Nakagawa et al. (2002) showed that nese with English abstract). in contrast to green algae, blue-green algae and diatoms are Abe, S., Iguchi, K., Nagumo, T., Katano, O. (2005). Distribution of well digested by ayu. ayu (Plecoglossus altivelis) and benthic algal flora in the Yaku Homeothrix varians has been reported in river reaches Island, Kagoshima Prefecture, Japan. Jpn J Phycol, 53, 1-6. (in downstream from dams in southwestern Japan (Morishita Japanese with English abstract). and Morishita, 2000). However, Abe et al. (2000) claimed Hirano, T., Suyama, M. (1983) Fatty acid composition and its sea- that the dominance of H. varians was due high grazing pres- sonal variation of lipids of wild and cultured ayu. Bulletin of the sure by ayu, because filamentous blue-green algae grows Japanese Society of Scientific Fisheries. 49, 1459-1464. (in Japa- faster than diatoms. Although we did not examine the rea- nese with English abstract). sons underlying the dominance of H. varians in the Yahagi Hirano, T., Zhang, C-H., Morishita, A., Suzuki, T., Shirai, T. (1992). River, differences the quality of the ayu food source does not Identification of volatile compounds in Ayu and its feeds. Nippon appear to explain the differences observed in sensory tests. Suisan Gakkaishi, 58, 547-557. Another reason for the low evaluation scores of the fish Hirose, H., Yamagishi, T. (eds.) (1977). “Illustrations of the Japa- from Sta. Y1 may be due to the effect of domestic sewage on nese fresh-water algae.” Uchida-roukakuho, Tokyo. (in Japanese). water quality. Ose et al. (1975) proposed that volatile fatty Hitomi, H. (1697). “Details of Food Material and Preparation in acids originating from domestic sewage confer an unfavor- Japan (Honcho Shokkan 3). Reprinted by Shimada I (2004).” able odor to fishes. It is of interest that the scores of aroma Heibon-sha, Tokyo. (in Japanese). agreed with the scores of general preference. From this point Hodoki, Y., Murakami, T., and Azuma, M. (2003). Relationship of view, the low score for the uppermost station (Y5) seems between body size and stomach content of adult Ayu in River odd and we suspect that the quality of water released from Kumagawa. Report of Nature Conservation Society of Japan, 94, dams is responsible; the largest dam on the river, the Yahagi 11-20. (in Japanese). Dam, is located in the upstream reaches of the river. The Jeong, B-Y., Moon, S-K., Jeong, W-G. (2000). Lipid classes and impounded water of is channeled and released immediately fatty acid compositions of wild and cultured sweet smelt Pleco- upstream of Y5. Loads of organic matter due to planktonic glossus altivelis muscles and eggs in Korea. Fisheries Science, production in reservoirs and lakes is most pronounced im- 66, 716-724. mediately downstream from such impoundments, becoming Josephson, D.B., Lindsay, R.C., Stuiber, D.A. (1983). Identification less severe as the water flows downstream (Murakami et of compounds characterizing the aroma of fresh whitefish (Core- al., 1994). Thus, compared to organic loads downstream, gonus clupeaformis). J. Agric Food Chem., 31, 326-330. the loads at Y5 are considered to be high. In addition, some Krammer, K. and Lange-Bertalot, H. (1991). “Bacillariophyceae planktonic blue-green algae behave as odorants (Oohashi and 3. Teil: Centrales, Fragilariaceae, Eunotiaceae.” Gustav Fisher, Hatama, 2005). The problematic odors in fishes are serious Stuttgart. in some Japanese rivers, although causal factors and contrib- Murakami, T., Isaji, C., Kuroda, N., Yoshida, K., Haga, H., Wata- uting processes are not clear (Yasuhara, 1995). Further inves- nabe, Y., Saijo, Y. (1994). Development of potamoplanktonic dia- tigations are therefore required to resolve this issue. toms in downreaches of Japanese rivers. The Japanese Journal of Sensory Analysis of the Taste of Ayu Sampled from Different River Environments 483

Limnology, 55, 13-21. Ose, Y., Funasaka, R., Sato, T. (1975). Studies on the offensive- Morishita, I. and Morishita, M. (2000). Ecology of Dam Impound- odor fish of the . Volatile Fatty acids as one of the ment. Environmental Conservation Engineering, 29, 906-911. offensive-odor substances. The Journal of Hygienic Chemistry, Nakagawa, H., Umino, T., Sekimoto, T., Ambas I., Montgomery, W., 21, 28-34. Nakano, T. (2002) Characterization of the digestive tract of wild Shiragane, A. (2002). Water quality of the Yahagi River - mainly at ayu. Fisheries Science, 68, 341-346. Fusso Waterfront Park. Report of Yahagi River Institute, 6, 99- Niimi, I. (1997). The present situation and issues on the Yahagi 111. (in Japanese). River and the details of the establishment of the Toyota Yahagi Takahashi, I. and Niimi, K. (1999). Life history of ayu, Plecoglos- River Institute. Report of Yahagi River Institute, 1, 1-6. (in Japa- sus altivelis, in the Yahagi River. Report of Yahagi River Institute, nese with English summary). 3, 247-267. (in Japanese with English summary). Oohashi, Y. and Hatama, T. (2005). Attached microalgal flora and Uchida, A. (2002). The contents of the digestive tract of ayu in the species composition in the gut content of ayu, Plecoglossus al- middle reach of the Yahagi River. Report of Yahagi river Institute, tivelis, from the lower side Ayu fishing area of the Abugawa Dam 6, 5-20. (in Japanese with English abstract). across the Abu River. Bull Yamaguchi Pref Fish Rec Ctr., 3, 137- Yasuhara, A. (1995). Odorous volatile components in Ayu. Toxico- 138. logical and Environmental Chemistry, 47, 197-201.