Original Communication

European Journal of Clinical Nutrition (2003) 57, 743–752. doi:10.1038/sj.ejcn.1601606

Glycemic index of single and mixed meal foods among common Japanese foods with white rice as a reference food

M Sugiyama1, A C Tang2, Y Wakaki3 and W Koyama3

  1. 1National Institute of Health and Nutrition, Japan
  2. 2Department of Health Promotion Sciences, School of Health Science and Nursing, Faculty of Medicine, The Tokyo University, Japan
  3. 3Kumamoto Health Management Center, The Japanese Red Cross Company, Japan

Correspondence: M Sugiyama, Division of Applied Nutrition, Laboratory of Nutrition Care and Management, 1-23-1, Toyama, Shinjuku-ku, Tokyo 162-8636, Japan. E-mail: sugiyama@hih.go.jp

Guarantor: M. Sugiyama.

Contributors: M Sugiyama provided the concept and design of the study. AC Tang carried out the literature research, data analyses and prepared the report. YW was responsible for data acquisition and analyses. WK was also responsible for the data acquisition.

Top

Abstract

Objective: The objectives were to examine the feasibility of using white rice as a reference food in the study of glycemic index (GI) and to examine the GI values of both single and mixed meal foods among rice species, processed rice products, beans, and dairy products.

Design: Subjects were served with 50 g carbohydrate content of white rice at least two times (maximum three times) and test food once after separate overnight fasts. Capillary blood glucose measurements were carried out before and during 120 min after each food load.

Setting: The study was carried out in an outpatient setting.

Subjects: A total of 58 (38 females and 20 males) nondiseased subjects, mean aged 37 y and mean BMI 22 kg/m2 were included.

Result: The correlation between incremental area under curve of white rice and glucose was r=0.853 (n=10, P<0.0001) and white rice was considered suitable to be used as a reference food. Among mixed meal foods, the combination of carbohydrate foods with vinegar, dairy products, and bean products significantly decreased the GI value of white rice of 20–40%. The reduction of GI occurred whether the foods were taken together, before or after rice intake. GI of noodles such as udon, soba, and spaghetti showed low GI values.

Conclusion: White rice could be used as a reference food in determining GI values of foods. A total of 32 single and mixed meal Japanese common food products were examined for their GI values.

Keywords:

glycemic index, rice, bean, dairy, vinegar, Japanese food

Top

Introduction

Jenkins (Jenkins et al, 1981) introduced the glycemic index (GI) in 1981 by comparing the postprandial blood glucose incremental area under curve (IAUC) of different carbohydrate foods. Since the introduction of the GI, numerous studies were conducted to test the GI of different foods and an International GI Table was published (Foster-Powell & Brand Miller, 1995). Among published literature, most studies were based in Western cultures and had used either white bread or glucose as the reference foods. For most Asian population who consume white rice as a staple food, rice is more readily available and is considered more palatable. For routine use, white rice used as a reference food would be more convenient in these countries when compared to white bread. Therefore, it is necessary to consider the use of white rice as a reference food. In addition, studies conducted among Japanese population with the internationally adopted GI evaluation method have not been reported. As a result, limited items published in the literature can be applied as common Japanese foods. Prior to the utilization of the GI concept in Japanese research and clinical setting, it is considered necessary to test the GI of commonly consumed Japanese foods among the local population with white rice as a reference. The objectives of the current study were to examine the feasibility of using white rice as a reference food and to test the GI of some common Japanese food.

Top

Method

Subject

A total of 58 healthy subjects (38 females and 20 males) who self-reported that they were not diagnosed with glucose intolerance or diabetes (based on their blood glucose and HbA1c examinations from the past 1 year) were included. The demographic characteristics of the subjects are shown in Table 1. These subjects had an average age of 37plusminus10 and BMI 22plusminus2.7 kg/m2 (meanplusminuss.d.). In all 25 subjects had family history of diabetes, 11 subjects took medication or supplements not related to glucose metabolism. Informed consent was obtained from each of the subjects according to the requirements set by the ethics committee of the Japanese Red Cross Company, Kumamoto Health Management Center, which conformed to the Helsinki Declaration.


Reference food

The reference food was the aseptic packed Satou Rice, which was labeled to contain 34 g of carbohydrate per 100 g of rice. Each portion of the cooked rice weighed 147 g and contained 50 g of carbohydrate value.

Test food

Test foods were chosen from common Japanese processed rice products, beans, and dairy foods. A total of 32 items were tested for their GI based on white rice as the reference food. Each test food contained 50 g of carbohydrate. In mixed meals or combination meals, the total nutrition value summed up to approximately 50 g carbohydrate. For example, for white rice and yogurt, 132 g of white rice (44.9 g of carbohydrate) and 100 g of yogurt (5.3 g carbohydrate) were given as one test food item. The reference food, test foods, nutrition component, and methods of offering foods are shown in Table 2.


Study protocol

The experiment was based on the method by Wolever (Wolever et al, 1991). Groups of 9–11 subjects were derived from the total pool of 58 subjects to test for one food. All subjects took 50 g carbohydrate load of reference food (white rice) at least two times (maximum three times) and test food once on random days. A subject who took more than one test food had at least 2 days of wash-out period in between food testing. Subjects started fasting from 09:00 on the previous evening before examination. Meals were given to the subjects from 07:00 to 09:00 on random days. Foods were consumed within 15 min. The subjects carried out self-monitoring blood glucose measurement seven times in total: at fast, and at 15, 30, 45, 60, 90, and 120 min after load. Capillary blood glucose measurement was performed using self-administered blood measuring device (Glutestase, Sanwa Chemical Research Institute Co. Inc.). The reliability of the device was tested in another occasion by measuring the same blood sample for 30 times and the coefficient of variation (CV) obtained was below 2.9. The correlation coefftcient of blood glucose value measured with self-administered blood measuring device and the value measured using the glucose automatic analysis device (The First Co. Inc. Kyoto Science GA-1 140) was r= 0.997 (n=62).

Calculation of the GI

The GI was defined as follows (Wolever, 1990):

Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

IAUC of each test food and mean IAUC for the initial two reference foods were calculated for each person. When the percentage change of the initial two reference foods IAUC was more than 25%, the subject took reference food the third time and mean IAUC of the closest two values was used as the reference. On the other hand, if the per cent change of the two closest IAUC exceeded 25%, the subject would be excluded from the analysis. Mean GI and s.d. of each test food was obtained from the remaining subjects in the group. Subjects who had GI exceeding 2s.d. were also excluded from the group and a final mean GI was recalculated to give each test food a GI.

Validity of the reference food and the conversion factor

To examine the feasibility of using white rice as a reference food, white rice was compared with 50 g pure glucose (150 g of glucose-tolerance-test glucose solution; Shimizu Seiyaku). Similar to the protocol stated above, subjects took white rice at least two times and glucose solution once. The mean IAUC of white rice and the IAUC of glucose were compared among a group of 10 subjects. Correlation coefficient of the two IAUC was determined. To allow comparisons across different studies based on glucose or white rice as a reference, a conversion factor to adjust the glucose-based GI (GIg) of any test food to white-rice-based GI (GIwr) and vice versa was determined. For the white-rice-based GI to glucose-based GI conversion factor, it was calculated as: l00/(mean GIg WR), where mean GIg WR represents the mean glucose-based GI of white rice of the 10 subjects. GIg WR of each subject was calculated as (mean incremental area of white rice/incremental area of glucose) times 100.

Statistical analysis

Statistical analysis was performed using the SPSS program (Version 9.0). The correlation coefficient was analyzed using Pearson's correlation coefficient. The significance of difference between reference and test food was tested by one sample t-test of the mean. Repeated measurements two-way ANOVA was used to assess heterogeneity in glucose response between foods. Data were shown as mean valuesplusminuss.d., unless otherwise stated. P<0.05 was considered significant.

Top

Result

The correlation between IAUC of rice and glucose was r=0. 853 (n=10, P<0.002) (Figure 1). The high correlation indicated that using rice as a reference food would be possible. White-rice-based GI (GIwr) was higher than glucose-based GI (GIg by a factor of 100/GIg WR=100/80=1.2, where 100 is the GI of glucose and 80 is the GIg of white rice (Table 3). Comparison of the white-rice-based study with other glucose-based studies could be done by proportionately adjusting the glucose-based GI to white-rice-based GI by a factor of 1.2, and vice versa by a factor of 0.8, so that the adjusted reference value is ascribed to be 100.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Correlation between IAUC of white rice and IAUC of glucose.

Full figure and legend (16K)


The results of 32 Japanese foods are listed in Table 4. Among single and mixed meals of common Japanese foods, vinegar products, dairy products, and bean products reduce the GI of rice when consumed together. For example, the use of vinegar in rice (sushi, GI=67) and vinegar-pickled foods (GI=75–77) decrease the GI of rice by about 20–35%. Dairy products also had low GI and GI-lowering effect. When white rice was taken with 100 ml of milk, the GI of rice was significantly reduced to 59. The intake of milk immediately before or after rice both significantly reduced the GI of rice to 67 and 68, respectively. The intake of yogurt before or after rice also significantly reduced the GI of rice. The significant GI lowering effect of bean and bean products (fermented soybean, GI=68; roasted and grounded soybean, GI=68; bean paste soup, GI=74) were also observed. Significant difference between IAUC of the reference food and the test food were seen for dairy products, bean products, udon, and spaghetti.


Top

Discussion

In the current study, Japanese staple food, white rice was examined for its feasibility as a reference food. In addition, using the International GI determination method, single and mixed meals of white rice, processed rice products, beans, and dairy products were compared for their incremental area under blood glucose curves and the GI of each food product was evaluated based on white rice as the reference food. The results showed that IAUC of white rice and glucose had high correlation and, therefore, was considered suitable to be used as a reference food. Moreover, it was determined that white-rice-based GI was higher than glucose-based GI by a factor of 1.2. And, white-rice-based GI could be adjusted to glucose-based GI by a factor of 0.8. In determining the GI of mixed meal foods, vinegar, dairy products, and bean products had GI lowering effect when consumed with rice.

White rice is the staple food in Japan. It is a self-sufficient grain product and is considered a cultural food. According to the Japanese National Nutrition Survey (Department of Health Promotion and Nutrition, The Ministry of Health, 1996), the average consumption of white rice is 161 g/day/person and rice is consumed two times or more per day. Rice accounts for about 29% of the total energy consumed by an adult and constitutes for 14% of the total protein energy consumed per day. On the other hand, total energy from other grain products constitute for only about 12% of the total energy consumed per day. A common Japanese meal usually consists of white rice with a main dish (protein-rich foods such as fish, meat, bean, or bean products) and one or two side-dishes (vegetables, sea algae, milk, or milk products, etc). GI study based on such meal structure is important in Japan and probably in other Asian cultures that consume rice as the staple food as well. In the previous studies, glucose and white bread were commonly used as reference foods for calculating GI. It was reported that white-bread-based GI were higher than glucose-based GI by a factor of 1.42 (GI of white bread=70; Glucose=l00), and that the white-bread-based GI was adjusted to glucose-based GI by 0.7 (Foster-Powell & Brand Miller, 1995). White bread was used instead of glucose as a reference food because it was considered more palatable and was thought to correlate to insulin secretion better than glucose (Bornet et al, 1987; Wolever et al, 1988). However, for most Asian population who consume white rice as a staple food, white bread as a reference food is considered less applicable for a number of reasons. First of all, white rice is more conveniently available than white bread. If white bread is to be used as a reference food, it must be baked according to the standard method each time a food is tested for its GI. Such test protocol can be difficult to achieve. Since white rice is readily available and can be easily prepared, it is easier for standardization. Secondly, flour used in Japan may originate from different wheat specie and when used for bread, the result would not be comparable to the results published in other countries. Thirdly, just as white bread is more palatable in Western countries, white rice is considered more palatable among the Japanese population. Lastly, in studying Japanese mixed meal and added ingredients on the effect of GI, it is more 'consumer-friendly' to use the commonly consumed white rice as a reference and compare other Japanese foods with it. For these reasons, examining white rice as a reference food was thought to be of good significance.

Rice of other species and processed rice products are common foods in Japan. Among the European and American countries, observations of GI values using different rice species and processed rice (steamed, boiled, stir-fried rice) have been reported (Foster-Powell & Brand Miller, 1995). The results showed that rice tested among diabetics and nondiabetics varied from GI=38 to 94 (Glucose=l00). The great difference in GI of rice was because of a number of reasons. Studies have reported that the variation of species (Miller et al, 1992), process condition (Brand et al, 1985, Traianedes & O'Dea, 1986), food structure such as particle size and shape (Jar-vi et al, 1994; Behall et al, 2000), starch content such as amylose vs amylopectin (Behall et al, 1989; Panlasigui et al, 1991) would change the GI of foods significantly. Amylose content in starch has an effect on the GI where low-amylose rice and instant rice were reported to have a higher GIg (mean GI=88–91), while parboiled rice (mean GIg=47), brown rice (mean GIg=55) and long-grain white rice (GIg=56) had a lower GI. The species of rice used for the reference food and for the test foods in the current study was the Koshihikari, a species favored by most Japanese. It is known to have a relatively low-amylose content (17% amylose). The GIg of white rice was 80, which was similar to the results of low-amylose white rice published in other countries. Glutinous rice in our study showed GI higher than that of white rice. We deduced that it was because of the very low-amylose content of the specie.

In the current study, it was found that dairy products significantly reduced the GI of white rice when consumed together, prior to or after a carbohydrate meal. The results are consistent with previous studies where dairy products decreased the GI of foods. Among previous studies, the GI of corn flake alone was classified as a high GI food but when milk was added to corn flake, the GI decreased (Miller et al, 1998). Similarly, in our study, milk and yogurt, both, caused the GI of rice to decrease, whether the dairy products were taken with, before or after rice ingestion. In addition, when cheese was added to curry, the GI also decreased. Fat and protein in the dairy products were probably responsible for the GI lowering effect by lengthening the gastric emptying and absorption process (Collier & O'Dea, 1983). Although dairy products seem to be low in GI and has a GI lowering effect in general, it has been criticized that these foods caused excessive secretion of insulin (Liljeberg & Bjorck, 2001). Insulin response was not measured in the current study and it was impossible for us to deduce the insulin response of the dairy products as well as other food products. Nevertheless, future study should consider the measurement of insulin in addition to glycemic response. With the excessive insulin response of some foods, careful selection of low GI foods should be taken to meet individual needs.

In the current study, the addition of vinegar and vinegared foods to white rice reduced the GI of white rice. The acetic acid in vinegar was thought to be responsible for the antihyperglycemic effect. The amount of acetic acid to be effective could be as low as that found in sushi (estimated to be about 0.2–1.5 g/100 g). The antihyperglycemic effect of vinegar is consistent with other studies performed earlier (Brighenti et al, 1995; Liljeberg & Bjorck, 1998). Although vinegar could lower GI vales, the mechanism has rarely been reported. Most studies accounted the mechanism to be due to a delay in gastric emptying. In animal studies, Fushimi (Fushimi et al, 2001) showed that acetic acid could activate gluconeogenesis and induce glycogenesis in the liver after a fasting state. It could also inhibit glycolysis in muscles. The mechanism of vinegar is yet to be confirmed with further studies in such respect.

Bean and bean products that were tested in the current study showed low GI values and had GI lowering effect when consumed with a carbohydrate meal. Previous studies showed that beans, lentils, and legumes had low GI values. The mechanism of the low GI of these food products seem to be because of the soluble fiber, antinutrients available in the foods (Wolever & Bourne, 1990).

The GI of spaghetti showed similar result with previous studies and the GI of Japanese noodles (udon and soba) were almost identical to spaghetti showing significant lower GI than white rice. It has been reported that structural properties would affect the glycemic response of pasta (Katri et al, 2002). Thus, udon, soba, and spaghetti may have similar GI values because of their similar structural properties.

In the current study, white bread showed GI=92 that coverts to GI=110 (92 times 1.2) when Glucose=100. In other studies, where white bread was tested, it had an average GI of about 70 when Glucose=100 (Foster-Powll & Brand Miller, 1995). Bread used for current testing was baked according to the published standard method. It is difficult to explain the large difference between these results. However, the difference in specie between Japanese-produced flour and European/American-produced flour could be a possible factor that caused the difference. The variety in results gave a good reason for us to use a different reference food other than bread for GI testing in Japan.

Numerous food studies as well as intervention studies have been conducted over the past decade to examine the benefits of low GI diets. Among most studies, low GI diets were reported to have positive effect in blood glucose control among diabetes patients (Fontvieille et al, 1992; Wolever et al, 1992; Brand Miller, 1994; Salmeron et al, 1997; Jenkins et al, 1998; Jar-vi et al, 1999; Luscombe et al, 1999) and in blood lipid lowering effect (Jenkins et al, 1987). Since white rice is a major high GI carbohydrate source, understanding the general diet structure among the rice-dominant populations is necessary. Moreover, efforts to select combinations of carbohydrate foods that could reduce postprandial glucose response could contribute to preventive and clinical health care. The current study should serve as an important reference in the future when using GI as a tool for diet education not only among Japanese population but also other Asian population, who consume rice as a staple food and have similar meal structure. In conclusion, white rice can be used as the reference food in determining GI of various Japanese foods. Among common Japanese foods, vinegar, dairy products, and bean products significantly decreased the GI of rice when ingested together, before, or after meal.

Top

References

  1. Behall KM, Scholfield DJ & Hallfrisch J (2000): The effect of particle size of whole grain flour on plasma glucose, insulin and TSH in human subjects. J. Am. Coll. Nutr. 18, 591–597.
  2. Behall KM, Scholfield DJ, Yuhaniak I & Canary J (1989): Diets containing high amylose vs amylopectin starch: effects on metabolic variables in human subjects. Am. J. Clin. Nutr. 49, 337–344. | PubMed |
  3. Bornet FRJ, Costagliola D, Rizkalla SW, Blayo A, Fontvieille AM, Haardt MJ, Letanoux M, Tchobroutsky G & Slama G (1987): Insulinemic and glycemic indexes of six starch-rich foods taken alone and in a mixed meal by type 2 diabetics. Am. J. Clin. Nutr. 45, 588–595. | PubMed | ISI | ChemPort |
  4. Brand JC, Nicholson PL, Thorbum AW & Truswell AS (1985): Food processing and the glycemic index. Am. J. Clin. Nutr. 42, 1192–1196. | PubMed | ChemPort |
  5. Brand Miller JC (1994): Importance of glycemic index in diabetes. Am. J. Clin. Nutr. 59, 747S–752S. | PubMed |
  6. Brighenti F, Castellani G, Benini L, Casiraghi MC, Leopardi E, Crovetti R & Testolin G (1995): Effect of neutralized and native vinegar on blood glucose and acetate responses to a mixed meal in healthy subjects. Eur. J. Clin. Nutr. 49, 242–247. | PubMed | ISI | ChemPort |
  7. Collier G & O'Dea K (1983): The effect of coingestion of fat on the glucose, insulin, and gastric inhibitory polypeptide responses to carbohydrate and protein. Am. J. Clin. Nutr. 37, 94l–944.
  8. Department of Health Promotion and Nutrition, The Ministry of Health (1996): The 1994 Japanese National Nutrition Survey. Daichi Publishing.
  9. Fontvieille AM, Rizkalla SW, Acosta M, Bornet FR & Slama G (1992): The use of low glycaemic index foods improves metabolic control of diabetic patients over five weeks. Diabetic Med. 9, 444–450.
  10. Foster-Powell & Brand Miller J (1995): International tables of glycemic index. Am. J. Clin. Nutr. 62, 871S–893S. | PubMed | ChemPort |
  11. Fushimi T, Tayama K, Fukaya M, Kitakoshi K, Nakai N, Tsukamoto Y & Sato Y (2001): Acetic acid feeding enhances glycogen repletion in liver and skeletal muscle of rats. J. Nutr. 131, 1973–1977. | PubMed | ISI | ChemPort |
  12. Jarvi AE, Kalstrom BE, Granfeldt YE, Bjorck IE, Asp NG & Vessby BO (1999): Improved glycemic control and lipid profile and normalized fibrinolytic activity on a low-glycemic index diet in type 2 diabetic patients (1999). Diabetes Care 22, 10–18. | Article | PubMed | ChemPort |
  13. Jarvi A, Karlstrom B, Granfeldt Y, Bjorck I, Vessby B & Asp NG (1994): The influence of food structure on postprandial metabolism in patients with non-insulin-dependent diabetes mellitus. Am. J. Clin. Nutr. 59, 794S.
  14. Jenkins DJA, Wolever TMS, Buckley G, LamKy, Giudici S, Kalmusky J, Jenkins AL, Patten RL, Bird J, Wong GS & Josse RG (1998): Low-glycemic-index starchy foods in the diabetic diet. Am. J. Clin. Nutr. 48, 248–254.
  15. Jenkins D, Wolever T & Kalmusky J (1987): Low-glycemic index diet in hyperlipidemia: use of traditional starchy foods. Am. J. Clin. Nutr. 46, 66–71. | PubMed | ISI | ChemPort |
  16. Jenkins DJA, Wolever TMS, Taylor RH, Barker H, Hasemein F, Baldwin JM, Bowling AC, Newman HC, Henkins AL & Goff DV (1981): Glycemic index of foods: a physiological basis for carbohydrate exchange. Am. J. Clin. Nutr. 34, 184–190.
  17. Katri SJ, Leo KN, Kirsi HL, Kaisa SP, Jens JH & Hannu MM (2002): Postprandial glucose, insulin, and incretin responses to grain products in healthy subjects. Am. J. Clin. Nutr. 75, 254–262. | PubMed |
  18. Liljeberg EH & Bjorck I (2001): Milk as a supplement to mixed meals may elevate postprandial insulinaemia. Eur. J. Clin. Nutr. 55, 994–999. | Article | PubMed |
  19. Liljeberg H & Bjorck I (1998): Delayed gastric emptying rate may explain improved glycaemia in healthy subjects to a starchy meal with added vinegar. Eur. J. Clin. Nutr. 52, 368–371. | Article | PubMed | ISI | ChemPort |
  20. Luscombe ND, Noakes M & Clifton PM (1999): Diets high and low in glycemic index versus high monounsaturated fat diets: effects on glucose and lipid metabolism in NIDDM. Eur. J. Clin Nutr. 53, 473–478. | Article | PubMed |
  21. Miller JB, Pang E & Bramall (1992): Rice—a high or low glycemic index food? Am. J. Clin. Nutr. 56, 1034–1036. | PubMed |
  22. Miller JB, Wolever TMS, Colagiuri S & Power KFB (1998): The Glucose Revolution. New York: Marlowe & Company, pp. 239–254.
  23. Panlasigui LN, Thompson LU, Juliano BO, Perez CM, Yiu SH & Greenberg GI (1991): Rice varieties with similar amylose content differ in starch digestibility and glycemic response in humans. Am. J. Clin. Nutr. 54, 87l–877.
  24. Salmeron J, Manson JE, Stampfer MJ, Colditz GA, Wing Al & Willett WC (1997): Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus in women. JAMA 277, 472–477. | Article | PubMed | ISI | ChemPort |
  25. Traianedes K & O'Dea K (1986): Commercial canning increases the digestibility of beans in vitro and postprandial metabolic responses to them in vivo. Am. J. Clin. Nutr. 44, 390–397.
  26. Wolever TMS, Boume GH (eds) (1990): The glycemic index. Aspects of some vitamins, minerals and enzymes in health and diseases. World Review Nutr Diet 62, 120–185.
  27. Wolever TMX, Jenkins DJA, Collier GR, Lee R, Wong GS & Josse RG (1988): Metabolic response to test meals containing different carbohydrate foods: relationship between rate of digestion and plasma insulin response. Nutr. Res. 8, 573–581.
  28. Wolever TMS, Jenkins DJA, Jenkins AL & Josse RG (1991): The glycemic index: methodology and clinical implications. Am. J. Clin. Nutr. 54, 846–854. | PubMed | ISI | ChemPort |
  29. Wolever TMS, Jenkins DJA, Vuksan V, Jenkins AL, Wong GS & Josse RG (1992): Beneficial effect of low-glycemic index diet in overweight NIDDM subjects. Diabetes Care 15, 562–564. | Article | PubMed | ChemPort |
Top

Acknowledgements

We thank the staff from the Japanese Red Cross Kumamoto Health Care Center and the staff from the Hospital Management Department, the National Institute of Health and Nutrition for their support and assistance. Also, we thank Dr N Hosoya (Professor Emeritus, the Tokyo University) for his invaluable advice to the research. The research was funded by the National Rice Association and the National Daily Promotion and Research Association of Japan.

Extra navigation

.

naturejobs

natureevents

ADVERTISEMENT