Astaxanthin added to the diet helps increase survival rate, growth and resistance to low dissolved oxygen stress of Japanese black tiger shrimp.
Natural carotenoids from astaxanthin-containing algae Haematococcus pluvialis (H) and astaxanthin-free algae Spirulina pacifica (S) and synthetic Carophyll Pink (A) astaxanthin were added to the diet at two concentrations of 50 (I) and 100 (II) mg kg −1 , resulting in seven pigmented diets HI, SI, AI, HII, SII, AII and HS (H-50 mg kg −1 +S-50 mg kg −1 ). A formulated diet without carotenoid supplementation was used as a control (C). Different diets were fed to Japanese giant tiger shrimp for 9 weeks. The effects of dietary carotenoids on survival, growth and pigmentation were compared by individual or combined treatments. A low dissolved oxygen test was conducted 2 weeks later and shrimp survival time and oxygen consumption rates were also compared between treatments. After 9 weeks of culture, shrimp fed control (C) had a significantly lower survival rate than shrimp fed pigmented feed. There was no difference in weight gain among all shrimp. C-fed shrimp had 66.4% less flesh astaxanthin (FA) and 75.5% less shell astaxanthin (SA) than shrimp fed a pigmented diet. Shrimp fed with I (AI, HI and SI) had 31.1% less FA and 29.6% less SA than shrimp fed with II (AII, HII, SII and HS). No significant differences were found when compared between other categories. The use of these three carotenoid sources for pigmentation in crustaceans is discussed along with carotenoid conversion, deposition, digestibility and absorption. When subjected to low dissolved oxygen stress, C-fed shrimp had a higher oxygen consumption rate (OCR) and shorter survival time (ST) than shrimp fed a pigmented diet. No differences in OCR or ST were found when compared across other categories.
1.INTRODUCE
Astaxanthin is the predominant carotenoid in penaeids (Ishikawa et al., 1966; Katayama et al., 1971, 1972; Okada et al., 1994). Because crustaceans cannot synthesize carotenoids de novo, astaxanthin or suitable precursors must be provided in the diet (Meyers and Latscha, 1997). Dietary supplementation of astaxanthin or its precursors has improved (Chien and Jeng, 1992; Liao et al., 1993) or modified (Mensaveta et al., 1993) the color of penaeid shrimp, especially are intensively raised animals, to get better market prices (Liao and Chien, 1994).
Several raw material sources such as the yeast Phaffia rhodozyma, the algae Dunaliella salina and the green algae Spirulina maxima, and synthetic hcarotene, canthaxanthin and astaxanthin have been used to pigment shrimp (Yamada et al., 1990; Chien and Jeng , 1992; Mensaveta et al., 1993; NegreSadargues et al., 1993; The effective variation of pigments from the above sources can be due to the type, composition and concentration of the pigments contained therein (Yamada et al., 1990; Howell and Matthews, 1991; Chien and Jeng, 1992; Menszveta et al., 1993; Petit et al., 1997), digestibility of the feedstock (Chien and Jeng, 1992), and possible presence of cofactors in the feedstock involved in the process absorption and deposition (Shahidi et al., 1998). The green unicellular freshwater alga, Haematococcus pluvialis, an alga with the potential to produce astaxanthin (Czygan, 1968; Borowitzka et al., 1991; Boussiba and Vonshak, 1991; Grung et al., 1992), has been proven to enhance the pigmentation of rainbow trout. (Sommer et al., 1991; 1992; Choubert and Heinrich, 1993), golden seabream (Gomes et al., 2002), and ornamental fish (Gouveia et al., 2003), but have never been used on armor to create pigment.
Besides the pigmentation properties of carotenoids, increasing attention is being directed toward determining the biological functions of astaxanthin in aquatic animals (Meyers and Latscha, 1997). Improvement in survival in Japanese tiger shrimp, Marsupenaeus japonicus (Chien and Jeng, 1992), and black tiger shrimp, Penaeus monodon (Thongrod et al., 1995), by dietary astaxanthin supplementation has been reported. . Recent studies have shown that increased resistance to hypoxic stress (Chien et al., 1999), salinity stress (Darachai et al., 1998; Mercchie et al., 1998; Chien et al. , 2003), heat stress (Chien et al., 2003). al., 2003), ammonia stress (Pan et al., 2003a), and pathological stress (Pan et al., 2003b) in penaeid postlarvae are associated with increased dietary and body astaxanthin.
Therefore, this study was designed to systematically compare different pigment sources (natural [H. pluvialis and Spirulina pacifica] with synthetic pigments [Carophyll Pink, Roche]; astaxanthin [H. pluvialis and Carophyll Pink] compared to non-astaxanthin [(S. pacifica)] and dietary levels, on their effects on astaxanthin in the body, survival and growth of Japanese tiger shrimp The effects on survival time and oxygen consumption rate in the low oxygen tension experiment were also evaluated.
2.Materials and methods
2.1. Experimental diet
Seven experimental diets were formulated and supplemented with carotenoids from three sources: algae H. pluvialis (H) and S. pacifica (S) and synthetic astaxanthin Carophyll Pink (A); each at two dietary carotenoid concentrations: 50 (I) and 100 (II) mg/kg (Table 1). A formulated diet without carotenoid supplementation was used as a control (C). Except for differences in carotenoid sources and concentrations, other ingredients in all eight diets were similar (Table 1). In the closely matched compositions obtained, the higher lipid content in the HI, HII and HS diets was attributed to the high lipid content in H. pluvialis, 20%.
2.2. Experiment on shrimp culture and pigmentation
Six-day-old Japanese tiger shrimp (M. japonicus) postlarvae were cultured in 20-ton tanks for 2 months until they reached an average weight of 0.4 g. They were then released into 24 outdoor tanks at a density of 40 fish/tank. The tank is black, plastic and round (200-l, 65 cm D65 cm H). To eliminate possible interference caused by shrimp ingesting pigments contained in algae, not only is the water used also filtered through a 1-Am filter and sterilized with ultraviolet light to remove most of the planktonic algae. but the tanks are also covered with black plastic to prevent algae growth. The bottom of the tank was covered with a 3 cm thick layer of beach sand to accommodate the burrowing behavior of kuruma shrimp. During the first 2 weeks of farming, all shrimp were fed a control diet to stabilize body pigmentation. A random sample of 3 shrimp from each tank was collected to determine initial weight and analyze body astaxanthin. The average initial astaxanthin concentrations on a dry weight basis were 161±32 mg/kg and 257±48 mg/kg for the meat and shell, respectively. Then, for nine weeks, experimental diets were fed to three tanks randomly assigned to each treatment. Shrimp were fed at a biomass rate of 5% in each tank/day, divided into 2 feedings at 09:00 and 18:00 h. Biomass was estimated based on data from weekly collection of all shrimp and daily mortality records. Aeration is provided at all times. Debris at the bottom of the tank, including uneaten food, is removed by siphoning daily and approximately one-third of the water is replaced daily to maintain proper water and sediment quality. Experimental water conditions are temperature 19–25o C, salinity 32–33‰, pH 8.2–8.3 and dissolved oxygen (DO) 5–7 m/L. Salinity was measured with a handheld refractometer (S/mill-E, ATAGO, Japan), pH with a pH meter (S30, MTANA, Schwerzenbach, Switzerland), and DO with a DO meter (YSI Model 59, Yellow Spring, Ohio, USA). At the end of the 9-week culture period, 2 shrimp were randomly sampled from each tank and immediately frozen in a 70°C refrigerator to analyze astaxanthin in their bodies.
2.3. Low dissolved oxygen tension experiment
This experiment was performed to investigate how differences in astaxanthin content in the body through dietary supplementation can affect survival time and oxygen consumption rate (oxygen consumed per unit animal weight per unit time) of shrimp under low dissolved oxygen (DO) conditions. Upon completion of the previous experiment, shrimp were fed the test and control diets for another 2 weeks. Four shrimp were randomly sampled from each tank. Two of them were immediately frozen for in vivo astaxanthin analysis. The remaining two types are used in the low DO pressure test. Each shrimp was first acclimated in oxygen-saturated seawater (temperature 25o C, salinity 32‰, pH 8.3, DO 4.8 mg/L, total ammonia-N 0.04 m/L) for a minimum of 4 hours in a 4-l container. A 300 ml bottle filled with seawater of similar quality is deoxygenated with bubbling nitrogen gas until DO is less than 0.5 mg/L
Then, a shrimp is placed in this bottle and the lid is tightly closed. Shrimp are considered dead when the gills do not move. Survival time was recorded. The DO in each bottle was determined when the shrimp was placed in the bottle and when the shrimp died. Background respiration was not measured and was considered negligible compared to shrimp respiration because the seawater used in this study was filtered through a 1-μm screen and sterilized with UV light.
2.4. Analysis of carotenoids and astaxanthin
The carotenoid content of the algae H. pluvialis and S. pacifica was analyzed. Samples were homogenized in a Waring blender for 5 min under nitrogen in the dark and extracted with acetone until no additional pigment was obtained. The extract was vacuum filtered through filter paper (5C, Toyo Roshi, Tokyo). Then n-hexane was added and the pigment was transferred to epiphase by adding distilled water. The epiphase was washed several times to remove acetone; Water was removed by adding 5–10 g Na2SO4 per 100 ml (Simpson et al., 1985; Chien and Jeng, 1992). The volume of the solution was recorded and the carotenoid concentration was measured using a spectrophotometer at an absorbance of 470 nm [E(1%, 1 cm)=2115] (Simpson et al., 1985).
Shrimp samples were weighed and dissected into shell (including carapace, telson and uropod) and meat (including viscera). Dissected sections were weighed, lyophilized, and weighed again to determine moisture content. Then, the dry sample was ground using a ceramic mortar and pestle and placed into a 50 ml polypropylene centrifuge tube. Then, 20 ml of acetone (0.05% butylated hydroxytoluene, BHT) was added as a solvent (Schwartz and Patroni-Killam, 1985; Barimalaa and Gordon, 1988), and the mixture was homogenized (Polytron PT-MR -3000) at 8000 rpm. in 1 minute. The contents of each tube were centrifuged (Hitachi 18 PR-52) under 4 8C at 12,700 ´ g for 15 min. The pellet was resuspended and centrifuged with an additional 20 ml of acetone until the acetone extract was clear. The pooled acetone extract was transferred to a 250 ml separating funnel, divided with 30 ml of n-hexane, and washed three times with 10% NaCl to remove excess acetone. The extract volume was reduced to 10 ml using a rotary evaporator (Rotavapor-R114 Waterbath-B480, Buˆchi, Switzerland) and then filtered through a 0.2-Am Millipore filter and stored in three colored 4-ml vials. brown. Astaxanthin was analyzed by high performance liquid chromatography (HPLC), using a Hitachi L-6200 pump, silica column (Lichrosorb Si-60 5 micro 2504.6 mm ID column, E. Merck Company), Hitachi L-4250 UV -VIS detector at 470 nm wavelength and integrated Hitachi D-2000 chromatograph. The operating conditions were: mobile phase, 14% acetone in nhexane; solvent flow rate, 1.5 ml/min; injection volume, 100 µl; and pump program, the sequence was 0–20 min Mixture A (acetone:n-hexane, 14:86) and 20.5–40 min Mixture B (100% n-heptane). . The system is controlled by a chromatographic data system (Scientific Information Services Corporation), which also integrates the areas under the peak. The standard was chromatographically pure astaxanthin (Hoffman La Roche Ltd., Basel, Switzerland).
2.5. Statistical analysis
One-way ANOVA was used to test the significance of the overall treatment effect on survival, body weight gain and astaxanthin of shrimp cultured for 9 weeks and survival time, oxygen consumption rate,
and astaxanthin in shrimp bodies were used in low DO stress experiments. Duncan’s multiple range tests (DMRTs) were performed for paired comparisons or individual (one-to-one) comparisons. In addition to DMRT, 7 orthogonal contrasts were performed for system or group (setto-set) comparisons: (1) No pigmentation (C) versus pigmentation (AI, HI, SI, AII, HII, SII and HS); (2) Supplement: 50 mg/kg (AI, HI, SI) vs. 100 mg/kg (AII, HII, SII and HS);
(3) Addition at 50 mg/kg: natural pigments (HI and SI) versus synthetic pigments (AI); (4) Natural: H. pluvialis (HI) vs. S. pacifica (SI); (5) Addition at 100 mg/kg: natural pigments (HII, SII and HS ) versus synthetic pigments (AII); (6) Natural: single algae (HII and SII) versus mixed algae (HS); and (7) Monospecific: H. pluvialis (HII) versus S. pacifica (SII). Linear regression analyzes were used on data from low DO pressure experiments to determine the relationship between survival time or oxygen consumption rate and astaxanthin content in shell and flesh samples. SAS software version 6.12 (SAS Institute, Inc., Cary, North Carolina, USA) was used for statistical analysis.
3. Result
3.1. Survival rate and weight gain
After 9 weeks of culture, the survival rate of shrimp fed C was significantly lower (pQ0.05) than shrimp fed other diets (Table 2). When compared overall using orthogonal contrasts (Table 3), the mean survival of shrimp fed with C (37%) was significantly lower than that of shrimp fed with the other experimental diets (54%). ). There were no significant differences in weight gain rates among all shrimp compared individually or collectively (Tables 2 and 3).
3.2. Astaxanthin content in the body
After 9 weeks of culture, astaxanthin in the meat of shrimp fed C was significantly lower than in shrimp fed the other experimental diets, except shrimp fed the SI diet (Table 2). Meat astaxanthin in shrimp fed with SI was less than shrimp fed with HII-, AII- and HS but did not differ from shrimp fed with SII. When compared overall (Table 3), C-fed shrimp had 66.4% less meat astaxanthin than the average meat astaxanthin of shrimp fed other diets. The average meat astaxanthin of shrimp fed the 50 mg/kg astaxanthin diet (AI, HI and SI) was 31.1% lower than that of shrimp fed the 100 mg/kg diet (AII, HII, SII and HS). No significant differences were found in other orthogonal contrasts.
The astaxanthin content in the shells of shrimp fed C was lower than the astaxanthin content in the shells of shrimp fed the other experimental diets. Shrimp fed with SI and HI had lower shell astaxanthin than shrimp fed with AII and HS. There was no difference in shell astaxanthin content between SI-, HI-, and AI-fed shrimp. When compared overall (Table 3), C-fed shrimp had 75.5% less shell astaxanthin content than the average shell astaxanthin content of shrimp fed the other experimental diets. The average shell astaxanthin content of shrimp fed the 50 mg/kg astaxanthin supplement (AI, HI and SI) was 29.6% lower than that of the shrimp fed the 100 mg/kg supplemented diet (AII, HII, SII and HS). No significant differences were found in other orthogonal contrasts.
3.3. Low dissolved oxygen tension
When subjected to low dissolved oxygen stress, Cfed shrimp had a higher oxygen consumption rate (OCR) and shorter lifespan than shrimp fed the other experimental diets (Tables 4 and 5). No significant differences were found in other individual or group comparisons.
4.Discussion
4.1. Astaxanthin content in the body
An increase in animal size, due to growth, dilutes carotenoid concentrations as weight and surface area increase faster than the rate of carotenoid absorption (Meyers and Latscha, 1997). This dilution effect is observed during development from ovary to protozoa I of Penaeus semiculcatus (Dall, 1995), during larval development (Mantiri et al., 1995) and at early postlarvae (stage IV) (McKay, 1987) of the European lobster Homarus gamarus, as well as in juveniles (Pan et al., 1999) and adult P. monodon (Mensaveta et al., 1993) . Assuming there are no carotenoids in the diet for absorption, the dilution factor will be the inverse function of the ratio between final weight and initial weight or surface area. In this study, shrimp raised with C had a weight gain of 281%. The carotenoid dilution factor in the body would be 26% (1/(100%+281%)), which can be considered quite close to the ratio between astaxanthin in the final and initial meat (34%) if the carotenoid present in diet C (3.2 mg/kg) is effectively assimilated.
Even when carotenoids are supplemented in the diet, the body’s original pigmentation is still difficult to maintain. In P.monodon, despite supplementing the diet with astaxanthin at 50 mg/kg (Mensaveta et al., 1993) or 80 mg/kg (Pan et al., 2001, 2003a), astaxanthin in the body still decreased. In this study, as long as shrimp were fed diets supplemented with carotenoids, 50 or 100 mg/kg, their final shell astaxanthin increased, ranging from 30% to 137%. However, meat astaxanthin content did not increase in shrimp fed diets supplemented with 50 mg/kg carotenoids, but increased by 8–24% in shrimp fed diets supplemented with 100 mg/kg carotenoids. Therefore, the levels of dietary carotenoids required to maintain astaxanthin in the body vary according to the target tissue. Comparing the increase in astaxanthin in peel (190%) compared with astaxanthin in meat (90%), it can be seen that astaxanthin deposition in peel is more sensitive than in meat to dietary carotenoid supplementation. It is expected that shrimp raised with synthetic astaxanthin (fed A) will have better pigmentation than shrimp raised with natural sources (fed H and S) at both levels of carotenoid supplementation because of the pigment source. in diet A there could be higher pigment deposition more effectively than in diets H and S.
Carophyll Pink in diet A, which contains mostly free astaxanthin, can accumulate directly in the tissues of kuruma shrimp (Yamada et al., 1990). However, in Hdiet, H. pluvialis contains a complex form of carotenoids in addition to esterified astaxanthin, including canthaxanthin, h-carotene, lutein and echinenone (Czygan, 1968; Choubert and Heinrich, 1993). These carotenoids need to undergo several biosynthetic conversions to astaxanthin (Simpson, 1982) before being deposited in shrimp tissue. In the S diet, S. pacifica contains mainly zeaxanthin and h-carotene (Soejima et al., 1980; Liao et al., 1993), which also require biosynthetic conversion before deposition. The pigmentation conversion of intermediate carotenoids in algae is slower than the direct deposition of synthetic astaxanthin. It has been found that unesterified astaxanthin is the most effective pigment for enhancing pigmentation in penaeid shrimp (Chien and Jeng, 1992; Negre-Sadargues et al., 1993; Mensaveta et al., 1993). Shahidi et al. (1998) concluded that crustacean pigmentation can be achieved more effectively when biosynthetic intermediates of substances structurally close to the storage form of the pigment are provided in their diet.
In salmon, esterified astaxanthin was found to be deposited less efficiently than unesterified astaxanthin (Storebakken et al., 1987). Choubert and Heinrich (1993) showed that cleavage of astaxanthin esters may be a limiting step for pigmentation, leading to lower carotenoid retention in rainbow trout by H. pluvialis compared to synthetic astaxanthin ( Sommer et al., 1991). Furthermore, the thick cell walls of algae can hinder digestion and thus the absorption of pigments by fish (Johnson and An, 1991). The pigmentation of rainbow trout fed with disrupted cells (cracked cell walls) of H. pluvialis was better than that of rainbow trout fed diets containing intact cells from H. pluvialis ( Sommer et al., 1991).
Similar results were obtained when disrupted and undisrupted Phaffia rhodozyma yeast were fed to Atlantic salmon (Tangeraas et al., 1989). Gouveia et al. (2003) suggested that the efficiency of staining ornamental fish Cyprinus carpio and Carassius auratus with Chlorella vulgaris (with thin cell membrane) is better than with H. pluvialis (with thick cyst wall) due to the high digestibility of ornamental fish. However, in this study, the results were not as expected. No differences in shell and flesh astaxanthin were observed in shrimp fed A, H and S.
This may be primarily due to sufficient time to convert unterminated carotenoids in H. pluvialis and S. pacifica to terminal or storage carotenoids. In crustaceans, the storage carotenoids are free astaxanthin, esterified astaxanthin and carotenoproteins (Castillo et al., 1982; Meyers and Latscha, 1997). For example, zeaxanthin, one of the major carotenoids in Spirulina sp., can be rapidly converted via 4-ketozeaxanthin to astaxanthin in P. monodon (Liao et al., 1993). Although pigmentation efficiency in crustaceans is largely related to the proximity of dietary carotenoids to astaxanthin in synthetic metabolism, the time required for each (step) of conversion has never been established or recorded.
Second, esterified astaxanthin can be accumulated in the tissues of crustaceans as effectively as non-esterified astaxanthin. This is evidenced by the dominance of esterified astaxanthin in epidermal tissue, as well as of complex carotenoproteins and carotenolipoproteins in the exoskeleton (Milicua et al., 1990; Meyers and Latscha, 1997).
Third, the digestibility of algal cell walls may not have significant adverse effects in crustaceans such as salmon (Sommer et al., 1991, 1992; Choubert and Heinrich, 1993), sea bream and sea bream. gold (Gouveia et al., 2002; Gomes et al., 2002) and ornamental fish (Gouveia et al., 2003). Algae are part of the natural diet of crustaceans and are the main source of nutrients during the larval stage (Smith et al., 1993). Finally, the high unsaturated fatty acid content of H. pluvialis and S. pacifica may favor the absorption of carotenoids. Carotenoids are fat-soluble pigments (Fox and Vevers, 1960), so their absorption and metabolism may also involve lipids. It has been observed that carotenoid biosynthesis or deposition can be adversely affected by poor lipid nutrition (Meyers and Latscha, 1997). In crustaceans, nonesterified and esterified carotenoids can accumulate as lipid dispersions in melanocytes (Fingerman, 1965), bound to chitin in the carapace (Ghidalia, 1985). Studies suggest that eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are the main fatty acids esterified with the chitin-linked astaxanthin fractions in shrimp shells (Meyers and Latscha, 1997). All the factors mentioned above may have caused complexity and variation in pigmentation and made the differences in shell and flesh astaxanthin between shrimp fed A, H and S insignificant.
4.2. Survival and weight gain
The survival rate of shrimp fed with C was lower than that of shrimp fed with pigmentation, demonstrating that 3.2 mg/kg of carotenoid diet was not enough to maintain shrimp survival. Similar survival rates between shrimp fed pigmented diets at different concentrations indicate that carotenoid supplementation at 100 mg kg/kg had no additional benefit on shrimp survival compared with 50mg/kg.
Yamada et al. (1990) also reported that M. japonicus fed a diet supplemented with astaxanthin (50~400 mg/kg) had a higher survival rate than those fed a control diet. Chien and Jeng (1992) also found a positive correlation between survival rate and optimal pigment concentration in shrimp tissue and suggested that pigmentation may play a role in improving shrimp survival. Borderner et al. (1986) on the other hand showed that regardless of concentration, as long as carotenoids were supplemented in the diet, crustacean survival was not affected. Overall, no positive effects were observed on the growth of penaeid shrimp fed diets supplemented with carotenoids (Yamada et al., 1990; Chien and Jeng, 1992; Negre-Sadargues et al. et al., 1993; Liao et al., 1993; Mensaveta et al., 1993; A similar result was obtained in this study. In the study of Thongrod et al. (1995), however, increased growth was observed in P. monodon fed at increasing levels (5 ~ 300 mg/kg) for 30 days. Petit et al. (1997) also found that dietary astaxanthin improved the growth rate and shortened the molting cycle of M. japonicus postlarvae during a 20-day culture period.
4.3. Low DO stress
C-fed shrimp were not only less tolerant to low DO stress but also consumed more oxygen under our conditions than shrimp raised with carotenoid-supplemented diets. Similar results were reported by Chien et al. (1999) on juvenile P. monodon fed a diet without added astaxanthin compared to those fed 360 mg kg 1 astaxanthin for 1 week. It is possible that oxygen-containing carotenoids such as astaxanthin and lutein (Soin, 1954; Czeczuga, 1979), in which oxygen is attached at the center of the hydrocarbon chain (Karnaukhov, 1979), may act as an intracellular oxygen reserve for respiratory process in the body. Oxygen stress in salmon eggs has been recognized (Craik, 1985).
Such intracellular oxygen storage (Ghidalia, 1985; Latscha, 1990; Oshima et al., 1993) or its equivalent electron acceptor (Ghidalia, 1985) may also allow crustaceans to survive under cold conditions. Hypoxic conditions are common in pond culture environments (Chien and Jeng, 1992). The high astaxanthin content of kuruma shrimp, about 90% of its total pigment (Ishikawa et al., 1966), may contribute to the shrimp’s ability to tolerate low DO and lower oxygen demand. This is evidenced by the fact that kuruma shrimp are one of the few species of penaeid shrimp that can be transported alive without water and the fact that during the day they burrow in sediments often in a hypoxic state. (Bailey-Brock and Moss, 1992).
In addition to functioning in the transport of calcium across membranes and serving as an oxygen reservoir in the neural respiratory chain, carotenoids also protect sensitive tissues and reactive compounds from oxidative damage. (Oshima et al., 1993). There have been several reports showing that astaxanthin functions as a potent antioxidant for shrimp against physical stressors (Darachai et al., 1998; Mercie et al., 1998; Chien et al., 2003). , chemistry (Pan et al., 2003a), and pathological stress (Merchie et al., 1998; Pan et al., 2003b).
Yew-Hu Chien, Wen-Chung Shiau