Postharvest Biology and Technology 24 (2002) 317– 325 www.elsevier.com/locate/postharvbio Biochemical changes during storage of sweet potato roots differing in dry matter content Zhitian Zhang a, Christopher C. Wheatley b, Harold Corke a,* a Department of Botany, Uni6ersity of Hong Kong, Pokfulam Road, Hong Kong, People’s Republic of China b International Center for Tropical Agriculture (CIAT), Cali, Colombia Received 25 July 2000; accepted 3 July 2001 Abstract Changes during storage were investigated in carbohydrate level, digestibility, a-amylase, trypsin inhibitor activity and pasting properties of roots of six genotypes of sweet potato (Ipomoea batatas (L.) Lam) differing in dry matter content. Most genotypes showed a slight decrease in starch content during 0 – 180 days of storage, but in the genotype Hi-dry, it decreased substantially. Alpha-amylase activity increased during the first 2 months of storage, followed by a decrease with continued storage to a level similar to that at harvest. The decline in starch content was correlated with a-amylase activity in the first 60 days storage (r= 0.80, P =0.06). Trypsin inhibitor activity (TIA) in the fresh roots varied among genotypes from 3.90 to 21.83 U/mg. Storage had little influence on TIA level. There was considerable genotypic variation in digestibility, with up to 27% reduction in digestibility after 120 days in storage. Glucose and sucrose concentration increased early in storage and then remained fairly constant. Storage reduced flour pasting viscosities, with up to nearly a 30% decline in peak viscosity. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Sweet potato; Postharvest storage; Trypsin inhibitor activity; Digestibility; Free sugars; a-amylase; Pasting viscosity 1. Introduction Sweet potato (Ipomoea batatas (L.) Lam) is grown throughout the tropics and subtropics, and ranks sixth or seventh among the most important food crops worldwide (Scott, 1992). China produces about 80% of the yearly global output, where it ranks fourth as a food crop, after rice, wheat and maize (Li et al., 1992). The trend in * Corresponding author. Tel.: + 852-2299-0314; fax: + 8522857-8521. E-mail address: hcorke@yahoo.com (H. Corke). utilization of sweet potato in China is shifting away from its use as a staple food to use it as a processed food, a raw material for industrial products, and for feed products. In rural areas, the limited transport infrastructure has encouraged small-scale local processing and the development of feed uses of sweet potato which are primarily household-based (Marter and Timmins, 1992). Although sweet potato is typically harvested in November or December, it is available for several months from storage in some areas in China. The storage of sweet potato induces many changes in the carbohydrate fraction of the roots. 0925-5214/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S0925-5214(01)00149-1 318 Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 The carbohydrate composition in sweet potato roots greatly affects the eating quality and processing traits (Picha, 1987; Walter and Palma, 1996). Generally, longer storage periods of raw roots prior to processing results in products with decreased firmness. Studies on the amylase activity in fresh and stored roots have been reported (Hagenimana et al., 1992, 1994; Takahata et al., 1995). Morrison et al. (1993) reported a marked difference in individual and total sugar concentrations among sweet potato lines, and significant varietal differences in aand b-amylase activity during storage. About 40% of China’s annual sweet potato production is destined for animal, particularly pig, feed use (Scott, 1992). Roots and leaves can be fed in many forms, but the low digestibility of raw starch is one of the constraints to feed efficiency. Dreher et al. (1984) reported that cereal starch is most susceptible to a-amylase digestion, while potato starch is resistant, and sweet potato starch has intermediate susceptibility. Significant variation in starch digestibility was observed among sweet potato genotypes (Noda et al., 1992; Zhang et al., 1995) and within the same genotype planted in different locations (AVRDC, 1988). Thus, selection for genotypes with high starch digestibility may be an effective way to increase sweet potato feed efficiency. Anti-nutritional factors such as trypsin inhibitors should also be taken into account. Trypsin inhibitors, proteinase inhibitors, which make proteins unavailable, are generally present in sweet potato roots. Variation in trypsin inhibitor activity among genotypes/cultivars has been reported (Dickey and Collins, 1984; Bradbury et al., 1985; Ravindran et al., 1995; Zhang et al., 1998). Prolonged storage of unprocessed sweet potato roots is important for food and feed availability. No comprehensive studies have been carried out on the effects of storage time on digestibility, starch pasting properties, and trypsin inhibitor activity of diverse genotypes. In this study, we report the diversity in these properties among six genotypes during storage. 2. Materials and methods 2.1. Samples Six sweet potato genotypes, Hi-dry, Yan1, Chao1, Yubeibai, Guang7 and Guang16 were used in this study. Hi-dry is a genotype from a germplasm collection supplied by the International Potato Center, Lima, Peru. It was developed by the USDA Vegetable Laboratory in Charleston, SC. Yan1 was developed in Shandong Province, China, and is mainly grown in Northern China, where it has wide adaptability. The other four genotypes were developed in Guangdong Province, China. Chao1 has been reputed to have favorable storage properties. These materials were grown in three replicates under experimental field conditions at the Kadoorie Agricultural Research Center of the University of Hong Kong in the New Territories, Hong Kong. After 173 days, all genotypes were physiologically mature. After harvest, five fresh roots selected from a pooled sample of the three replicates for each genotype were analyzed and fifteen roots were stored at 20 °C and 75% relative humidity. Roots were removed from storage at day 60, 120 and 180 for analysis. Only sound roots of 150– 250 g/root size were used. At each storage time, sweet potato roots were washed thoroughly, peeled, sliced into thin chips (1– 1.5 mm) and dried at 40 °C for 3 h, additionally dried in a freeze-dryer, then processed into 100-mesh flour (SPF) for analysis. 2.2. Analysis Total starch content was determined using 100 mg samples by the amyloglucosidase/a-amylase method with the Megazyme Total Starch Assay Kit (Megazyme Ltd, Bray, Ireland) (McCleary et al., 1997). Alpha-amylase activity was determined using 500 mg samples by the Ceralpha a-amylase assay procedure (Megazyme Ltd, Bray, Ireland) (McCleary and Sheehan, 1987). Analysis for both tests was carried out on SPF. Trypsin inhibitor activity (TIA) measurement was based on method 71-10 (AACC, 1995). SPF (200 mg) was extracted in 25 ml 0.01 N NaOH for 3 h. The extraction solution was diluted four Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 319 Table 1 Total starch, a-amylase activity and trypsin inhibitor activity in fresh sweet potato roots at harvest Genotype Dry matter (%, Total starch (%, a-amylase activity (Ceralpha unit/g, dry basis) dry basis) dry basis) Trypsin inhibitor activity (U/mg, dry basis) Hi-dry Yan1 Chao1 Yubeibai Guang7 Guang16 Mean LSD (0.05) 33.5 9 0.9 29.3 9 1.6 22.6 9 0.6 27.9 9 0.1 26.9 9 1.2 24.3 9 0.4 27.4 2.59 16.5 91.84 18.6 92.56 3.90 90.18 4.99 90.17 8.74 90.89 21.8 91.74 12.41 3.50 73.6 9 0.5 55.3 9 0.1 46.8 9 2.0 52.6 9 1.1 57.6 9 3.4 49.6 9 1.1 55.9 4.8 0.41 9 0.01 0.81 9 0.01 1.73 9 0.06 1.25 9 0.18 1.14 9 0.04 1.44 9 0.04 1.13 0.20 Values are means of two replicates. times with distilled water before TIA determination. Flour digestibility was measured by the Gates and Sandstedt method as reported in Zhang et al. (1995) with some modification. SPF (500 mg) was placed in a weighed centrifuge tube (28.7× 104.2 mm, Nalgene Co, NY) with 15 ml phosphate buffer (0.15 M, pH 6.5), 30 mg CaCl2, 30 mg gelatin and 30 mg pancreatin (Sigma Chemical Co, St Louis, MO). The capped tubes were shaken at 37 °C at sufficient speed to keep the flour in suspension for 12 h, then 5 ml of 0.1 M H2SO4 was added to stop the reaction. The suspension was centrifuged for 10 min at 20000×g. The supernatant was decanted, and the residue pellet was dispersed with 15 ml of 80% ethanol, and recentrifuged for 5 min. The supernatant was decanted, and the tubes with the residue pellet were dried at 50 °C for 6 h, then at 80 °C to constant weight, cooled, and weighed. Flour digestibility was expressed as percent weight loss after digestion. A blank (without pancreatin) was included for each sample to adjust the results. To measure free sugar content, SPF (100 mg) was extracted with 10 ml 80% ethanol by shaking for 3 h at 37 °C. The solution was filtered through Whatman No. 42 filter paper and diluted 10 times with distilled water before HPLC analysis. A DX 500 HPLC system (Dionex Co, Sunnyvale, CA) was used to determine glucose, sucrose and fructose concentration in the extraction solution. The system included an ED 40 electrochemical detector, a 4 × 250 nm Carbopac™ Pa1 column (Dionex) and an LC30 stainless steel automated Rheodyne injection valve with a 25 ml fixed loop. PeakNet software and a DX LAN™ computer interface card (Dionex Co) were used. The mobile phase consisted of 200 mM NaOH-distilled water (2:98, v/v). The flow rate was 1.0 ml/min. The column was kept at room temperature (20– 22 °C). External standard solutions of glucose, sucrose and fructose (Sigma Chemical Co, St Louis, MO) were used for identification and quantification. A Rapid Visco-Analyzer 3-D (RVA) (Newport Scientific Pty Ltd, Warriewood, Australia) was employed to determine the SPF pasting properties. SPF (3.5 g, 14% m.b.) and 24.5 ml of 1.0 mM AgNO3 were mixed in the aluminum RVA sample canister. A programmed 22.0 min heating and cooling cycle was used. The sample was held at 50 °C for 1 min, heated to 95 °C in 7.5 min, held at 95 °C for 5.0 min, cooled to 50 °C in 7.5 min, and then held at 50 °C for 1 min. Parameters recorded from the pasting curve were peak viscosity (PV, the highest viscosity of the paste after gelatinization), hot paste (holding) viscosity (HPV, the viscosity of the paste after shear thinning), and cool paste viscosity (CPV, the final viscosity at 50 °C). Breakdown (BD) was computed as PV-HPV, and setback (SB) as CPV-HPV. Data analysis was carried out using the Statistical Analysis System version 6.1 for Windows (SAS Institute, Cary, NC). Comparison of means was carried out using the least significant difference (LSD) test. Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 320 Table 2 Digestibility of flour of sweet potato roots during storage Genotype Table 3 Changes in free sugar concentration in sweet potato roots during storage Digestibility (% weight loss after digestion) Genotype At harvest Hi-dry Yan1 Chao1 Yubeibai Guang7 Guang16 Mean LSD (0.05) 59.2 9 4.0 57.0 9 1.8 47.7 9 3.5 53.7 9 2.5 55.4 9 4.9 51.2 9 3.3 54.0 8.8 60 days storage 120 days storage 48.9 93.1 51.8 94.1 45.2 92.5 52.7 93.4 55.7 95.1 40.8 91.8 49.2 9.5 32.5 9 2.4 46.8 9 2.5 42.6 9 4.1 45.8 9 3.3 43.2 9 2.0 35.0 9 3.4 41.0 8.1 Digestibility was expressed as percent weight loss after digestion. Values were means of two replicates. 3. Results and discussion 3.1. Total starch content Starch was the major component of sweet potato root dry weight, with a mean in the six genotypes of 55.9% at harvest. A wide variation in starch content was observed, with Hi-dry containing the highest starch content (73.6%), and Chao1 the lowest (46.8%) (Table 1). Dry matter and starch content of fresh roots was positively correlated (r=0.92, P B0.01). During storage, five of the six genotypes showed slight decreases in starch content, while Hi-dry demonstrated a dramatic decrease (Fig. 1A). Although Hi-dry showed the highest starch content at the time of harvest, it decreased to the third highest among the six genotypes after 6 months. Therefore, in screening for genotypes with high dry matter and starch content, the stability of these properties in unprocessed root over time should be taken into account. 3.2. Alpha-amylase acti6ity The average a-amylase activity at harvest in the six genotypes was 1.13 Ceralpha unit/g. This was generally comparable with the lower range of values for 44 Philippine genotypes reported by Collado et al. (1997). There was a four-fold difference among the genotypes, from Chao1 at 1.73 Sugar content (% dry matter) Harvest 60 days storage 120 days storage Hi-dry Glucose (G) Fructose (F) Sucrose (S) G+F+S 4.11 9 0.25 0.26 9 0.04 0.42 9 0.07 4.79 9 0.23 6.23 9 0.47 0.47 9 0.07 0.58 9 0.07 7.28 9 0.47 5.20 9 0.43 0.72 9 0.11 0.64 9 0.14 6.56 9 0.17 Yan1 G F S G+F+S 3.459 0.64 3.07 9 0.24 3.14 9 0.08 9.66 9 0.79 4.89 9 0.30 2.859 0.28 3.49 9 0.19 11.2 9 0.39 3.94 9 0.08 4.12 9 0.12 4.29 9 0.44 12.4 9 0.24 Chao1 G F S G+F+S 3.809 0.49 4.67 9 0.28 4.02 9 0.05 12.59 0.17 5.63 9 0.23 4.579 0.66 5.299 0.25 15.5 9 0.65 6.10 9 0.13 4.33 9 0.24 5.19 9 0.15 15.6 9 0.26 Yubeibai G F S G+F+S 4.379 0.20 2.169 0.38 2.33 90.24 8.9 9 0.43 5.89 9 0.20 2.20 9 0.18 3.05 9 0.10 11.19 0.28 5.83 9 0.27 2.24 9 0.15 3.17 9 0.23 11.39 0.19 Guang7 G F S G+F+S 4.10 9 0.13 3.89 9 0.16 3.38 9 0.03 11.4 9 0.06 6.02 9 0.12 2.519 0.14 4.049 0.11 12.6 9 0.16 5.73 9 0.55 2.31 9 0.14 3.84 90.35 11.9 9 0.76 Guang16 G F S G+F+S 2.38 9 0.11 4.76 9 0.24 3.40 9 0.35 10.5 9 0.47 3.999 0.35 4.72 90.12 4.80 9 0.13 13.5 9 0.33 4.14 9 0.22 4.909 0.29 5.55 90.34 14.69 0.17 Sugar concentration was expressed on a percent dry matter basis. Values were means of three replicates. Ceralpha units/g, to Hi-dry at 0.41 Ceralpha units/g (Table 1). Generally, a-amylase activity was low at harvest, increased to higher levels after 60 days of storage, and gradually declined to harvest levels after 180 days storage (Fig. 1B). The extent of change in a-amylase activity during storage varied with genotypes. Yan1 was the most stable, while Hi-dry showed the largest change in a-amylase activity. Although Hi-dry had the lowest a-amylase activity at harvest, the level in- Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 321 Fig. 1. Changes in (A) total starch content (LSD = 5.0), (B) a-amylase activity (LSD = 0.3), (C) trypsin inhibitor activity (TIA) (LSD= 4.0), in sweet potato roots of six genotypes during storage. LSD values are expressed at PB 0.05 level. 322 Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 creased 10-fold after 60 days (Fig. 1B). Increase in a-amylase activity was correlated (r=0.80, P = 0.06) with the reduction in starch content in the stored roots. Yan1 had the smallest change in a-amylase activity and had the most stable starch content. Hi-dry had the largest increase in a-amy- lase activity, and the largest reduction in starch content during the first 60 days storage. Hagenimana et al. (1992) reported that in an in vitro experiment a-amylase was unable to hydrolyze native sweet potato starch granules. They postulated that starch phosphorylase was vital for the Fig. 2. RVA pasting profiles (in 1.0 mM AgNO3 solution) of flour from sweet potato roots: (A) averaged over six genotypes, showing mean profile at harvest and mean profile after 60 days storage. (B) Profile at harvest and after 60 days storage for genotypes Hi-dry and Yubeibai. Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 323 Table 4 Pasting properties of sweet potato flour of roots during storage (RVU, Rapid Viscosity Units) Genotype Hi-dry Yan1 Chao1 Yubeibai Guang7 Guang16 Mean Peak viscosity (RVU) Hot paste viscosity (RVU) Cool paste viscosity (RVU) Harvest 60 days storage Change (%) Harvest 60 days storage Change (%) Harvest 60 days storage Change (%) 568 373 268 368 425 306 385 395 301 248 242 326 304 303 −30 −19 −7 −34 −23 −1 −21 230 243 186 207 201 197 211 187 221 178 189 181 174 188 −19 −9 −4 −9 −10 −12 −11 335 340 256 299 289 284 301 285 321 264 264 257 252 274 −15 −6 +3 −12 −11 −11 −9 in vivo degradation of raw sweet potato starch during sprouting. However, Takahata et al. (1995) reported that the dextrins produced during storage had a close relationship with a-amylase activity in each line, suggesting that a-amylase in sweet potato roots does play a key role in starch degradation during storage. 3.3. Trypsin inhibitor acti6ity Substantial trypsin inhibitor activity (TIA) in fresh roots was found. The mean of the six genotypes was 12.4 U/mg. Values ranged from 3.9 U/mg in Chao1 to 21.8 U/mg in Guang16 (Table 1). Both qualitative and quantitative differences in trypsin inhibitor proteins in sweet potato roots may be responsible for the genotypic variation in TIA. Dickey and Collins (1984) studied sweet potato roots of four cultivars and found seven different trypsin inhibitor bands after electrophoresis at pH 8.9 in 7.5% acrylamide gels. Wang and Yeh (1996) reported that two forms of trypsin inhibitor (Mr =31 and 21 kDa) were present in sweet potato roots of five cultivars, and the TIA varied among cultivars. Measurement of TIA in roots stored over time showed that the six genotypes maintained largely constant levels, with only slight decreases in TIA in Hi-dry and Yan1 at the late storage stage (Fig. 1C). Thus, storage had little influence on the TIA in the genotypes involved in this study. 3.4. Flour digestibility It has been demonstrated that digestibility of sweet potato flour is highly correlated with starch digestibility (Zhang et al., 1993). The mean digestibility of flour of fresh roots of the six sweet potato genotypes was 54.0%, with a range from 47.7% in Chao1 to 59.2% in Hi-dry (Table 2). These results were consistent with earlier studies (AVRDC, 1988). Digestibility of flour tended to decrease with storage (Table 2). The mean digestibility of the six genotypes decreased from 54.0% at harvest time to 49.2% after 60 days, to 41.0% after 120 days in storage. However, there was considerable variation in the change of digestibility among different sweet potato genotypes. Yan1 and Chao1 showed relatively constant digestibility during storage. A large change in digestibility was observed in Hi-dry (decrease of 26.7% after 120 days). Although Hi-dry showed the highest digestibility at harvest time, it had the lowest value after 120 days storage. 3.5. Free sugar content The levels of glucose, fructose and sucrose in fresh roots varied among genotypes. The mean total of these sugars was 9.6%, ranging from 4.8% in Hi-dry to 12.5% in Chao1 (Table 3). These values were consistent with earlier findings (Martin and Deshpande, 1985; Truong et al., 1986; 324 Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 Babu, 1994). In addition, the sugar composition was genotypically variable, e.g. glucose accounted for 86% of the three sugars in Hi-dry, but 22 – 49% in the others. The level of glucose varied among genotypes from 2.4 to 4.4%. Hidry was particularly low in fructose and sucrose and hence total sugar. Five of the six genotypes showed increased total sugars in the earlier stages of storage and maintained relatively constant levels with further storage. However, Guang7 showed a negligible change in total sugars during the entire storage period. These results were somewhat different from those of Morrison et al. (1993), who suggested that changes in individual and total sugar concentration for sweet potato lines during storage were relatively minor. 3.6. SPF pasting properties To eliminate the influence of amylase on the pasting process, the RVA viscoamylography of the SPF was determined in an amylase inhibitor solution (1.0 mM AgNO3). The pasting parameters of the RVA pasting profiles of SPF of fresh roots (Table 3) showed that mean PV, HPV and CPV at harvest were 385, 211 and 301 RVU, respectively. There was a wide variation in pasting properties among the six genotypes, Hi-dry had the highest PV (568 RVU), and Chao1 the lowest (268 RVU). Chao1 also had the lowest HPV (186 RVU) and CPV (256 RVU). Genotypic variation in PV was greater than HPV and CPV. RVA pasting profiles changed after storage for 60 days. Considering the average pasting profile of the six genotypes, storage had a delaying effect on gelatinization, and PV was reached at a higher temperature (Fig. 2A). Storage resulted in reduced pasting viscosities. However, changes in the pasting profile varied considerably among genotypes. Chao1 and Guang16 showed smaller changes in the pasting properties, and intermediate changes were observed in Yangshu1 and Guang7 (Table 4). Hi-dry and Yubeibai had sharp declines in RVA pasting viscosities (Fig. 2B). Of the pasting parameters, PV was affected more than HPV and CPV by storage. In Hi-dry, PV fell from 568 RVU at harvest to 395 RVU after 60 days storage, about a 30% reduction (Table 4). Since starch is the main component of sweet potato roots, the decrease in starch content during storage would contribute to the reduced pasting viscosities. Hidry, with the largest decrease in starch content, showed the greatest reduction in all pasting parameters. However, this trend did not occur in all genotypes. Although Yubeibai maintained a constant starch content in storage (Fig. 1A), it had a significant reduction in PV after 60 days storage (Table 4). This indicated that changes of other components also contributed to the changes in pasting properties. Effects of sugars on starch pasting properties have been previously reported. Bean and Yamazaki (1978) showed that sucrose delayed starch gelatinization. Deffenbaugh and Walker (1989) showed that addition of sugar to starch tended to increase peak viscosity, but peak viscosity decreased at high sugar concentration. 4. Conclusions With increasing use of sweet potato flour and purified starch for added-value processed products in China, control of raw material quality is becoming more critical. Genotype and growing environment are well-known to affect physical qualities of sweet potato, particularly of parameters related to starch viscosity. Genotypic differences in postharvest changes in quality parameters have also been shown to be important. Knowledge of these changes can enable development of models to plan the optimum storage and processing time for different genotypes to match specific industry needs. Acknowledgements Financial support was received from the Hong Kong Research Grants Council, the University of Hong Kong Committee on Research and Conference Grants, and the International Potato Center (Lima, Peru). Z. Zhang et al. / Posthar6est Biology and Technology 24 (2002) 317–325 References AACC, 1995. Approved Methods of the American Association of Cereal Chemists. 9th edn. Method 71-10 approved November 1973, reviewed October 1982. The Association: St. Paul, MN. AVRDC, 1988. Chemistry-screening for starch digestibility of sweetpotato. Progress report-1988. Asian Vegetable Research and Development Center (AVRDC), Taiwan, pp. 249 – 253. Babu, L., 1994. Changes in carbohydrate fractions of sweet potato tubers on processing. Trop. Agric. 71, 71 – 73. Bean, M.M., Yamazaki, W.T., 1978. Wheat starch gelatinization in sugar solutions. I. Sucrose: microscopy and viscosity effects. Cereal Chem. 55, 936 – 944. Bradbury, J.H., Hammer, B., Nguyen, T., Anders, M., Millar, J.S., 1985. Protein quantity and quality and trypsin inhibitor content of sweet potato cultivars from the highlands of Papua New Guinea. J. Agric. Food Chem. 33, 281 – 285. Collado, L.S., Mabesa, L.B., Corke, H., 1997. Genetic variation in color of sweetpotato flour related to its use in wheat-based composite flour products. Cereal Chem. 74, 681 – 686. Deffenbaugh, L.B., Walker, C.E., 1989. Use of the RapidVisco-Analyzer to measure starch pasting properties. I. Effect of sugars. Starch/Staerke 41, 461 – 467. Dickey, L.F., Collins, W.W., 1984. Cultivar differences in trypsin inhibitors of sweet potato roots. J. Am. Soc. Hort. Sci. 109, 750 – 754. Dreher, M.L., Dreher, C.J., Berry, J.W., 1984. Starch digestibility of food: a nutritional perspective. CRC Crit. Rev. Food Sci. Nutr. 21, 47 – 71. Martin, F.W., Deshpande, S.N., 1985. Sugars and starches in a non-sweet sweet potato compared to those of conventional cultivars. J. Agric. Univ. Puerto Rico 69, 401 – 406. Hagenimana, V., Vezina, L.P., Simard, R.E., 1992. Distribution of amylase in sweet potato (Ipomea batatas L.) root tissue. J. Agric. Food Chem. 40, 1777 – 1783. Hagenimana, V., Vezina, L.P., Simard, R.E., 1994. Amylolytic activity in germinating sweet potato (Ipomea batatas L.). J. Am. Soc. Hort. Sci. 119, 313 – 320. Li, W.G., Wu, X., Cai, H., Du, R., 1992. Sweet potato in China. In: Scott, G.J., Wiersema, S., Ferguson, P.I. (Eds.), Product Development for Roots and Tuber Crops, vol. 1, Asia. International Potato Center (CIP), Lima, Peru, pp. 41 – 50. Marter, A.D., Timmins, W.H., 1992. Small-scale processing of sweet potato in Sichuan province, Peoples’ Republic of 325 China. Trop. Sci. 32, 241 – 250. McCleary, B.V., Sheehan, H., 1987. Measurement of cereal a-amylase: a new assay procedure. J. Cereal Sci. 6, 237 – 251. McCleary, B.V., Gibson, T.S, Mugford, D.C., 1997. Measurement of total starch in cereal products by amyloglucosidase-a-amylase method. J. Assoc. Off. Anal. Chem. 80, 571 – 579. Morrison, T., Pressey, R., Kays, S.J., 1993. Changes in a-and b-amylase during storage of sweet potato lines with varying starch hydrolysis potential. J. Am. Soc. Hort. Sci. 118, 236 – 242. Noda, T., Takahata, Y., Nagata, T., Monma, M., 1992. Digestibility of sweet potato raw starches by glucoamylase. Starch/Staerke 44, 32 – 35. Picha, D.H., 1987. Carbohydrate changes in sweet potatoes during curing and storage. J. Am. Soc. Hort. Sci. 112, 89 – 92. Ravindran, V., Ravindran, G., Sivakanesan, R., Rajaguru, S.B., 1995. Biochemical and nutritional assessment of tubers from 16 cultivars of sweet potato (Ipomoea batatas L.). J. Agric. Food Chem. 43, 2646 – 2651. Scott, G.J., 1992. Transforming traditional food crops: product development for roots and tubers. In: Scott, G.J., Wiersema, S., Ferguson, P.I. (Eds.), Product Development for Roots and Tuber Crops, vol. 1, Asia. International Potato Center (CIP), Lima, Peru, pp. 3 – 20. Takahata, Y., Noda, T., Sato, T., 1995. Changes in carbohydrates and enzyme activities of sweet potato lines during storage. J. Agric. Food Chem. 43, 1923 – 1928. Truong, V.D., Christopher, J.B., Judith, A.M., 1986. Simple sugars, oligosaccharides, and starch concentrations in raw and cooked sweet potato. J. Agric. Food Chem. 34, 421 – 425. Walter, W.M., Palma, C.S., 1996. Effect of long-term storage on cell wall neutral sugars and galacturonic acid of two sweet potato cultivars. J. Agric. Food Chem. 44, 278 – 281. Wang, H.Y., Yeh, K.W., 1996. Cultivar differences in trypsin inhibitory activities of sweet potato leaves and tuberous roots. Taiwania 41, 27 – 34. Zhang, D.P., Collins, W.W., Belding, S., 1993. Improving sweet potato starch digestibility for animal feeding. HortScience 28, 325 –326. Zhang, D.P., Collins, W.W., Andrade, M., 1995. Estimation of genetic variance of starch digestibility in sweet potato. HortScience 30, 348 –349. Zhang, D.P., Collins, W.W., Andrade, M., 1998. Genotype and fertilization effects on trypsin inhibitor activity in sweet potato. HortScience 33, 225 –228.