J. Sci. Food Agric.

1981,32, 1021-1026

Starch and Sugar Transformation During the Ripening of Plantains and Bananas
John Marriott, Michael Robinson and Simon K. Karikaria
Tropical Prodircts Institute, 56 Grays Inn Road, London W C l X 8LU and aUniversity of Ghmia, PO Box 68, Legon, Accra, Ghana (Manuscript received 14 November 1980)

Ripening changes were studied in plantains of three cultivars, two Horn types and one French type, and compared with those in bananas ripened under the same conditions. Bananas contained 1 % starch when fully ripe and none when overripe, whereas plantains contained about 9% starch when fully ripe and 3 % when overripe (composition expressed as percentage fresh weight). Total sugar content was 23 % in fully ripe and overripe bananas but in plantains it increased from 20% when fully ripe to 27% when overripe. The ratio of glucose: fructose was approximately unity for bananas and plantains at all stages of ripeness. Sucrose comprised more than 70% of the total sugars in fully ripe bananas and plantains and about half of the total sugars in overripe fruits.

1. Introduction

Plantain fruits (Mum AAB group, plantain sub-group) are widely grown throughout the tropics. Plantains are nearly always cooked before use and may be boiled or dried either intact or after grating or pounding. This diversity of culinary processes depends on the compositional and textural changes which occur during ripening from a hard, essentially starchy, green fruit to a soft, sweet, yellow one. Even when ripe, plantains are firmer and more starchy than the types of dessert bananas used in major export trades (Musa, AAA group, Cavendish sub-group).l The compositional changes during ripening of exported bananas are well understood2 and the ripening process is controlled at terminal markets to facilitate marketing as dessert fruit.3 In exported cultivars of bananas, starch forms about 20%of the fresh pulp of the unripe, green fruit. It is almost completely hydrolysed during ripening with formation of sucrose, glucose, and f r u ~ t o s eA . ~more recent study on carbohydrate transformation in bananas, also detected traces of a trisaccharide.5 Starch hydrolysis in plantains is supposed to be slower and less complete than in bananas but the starch content of fully ripe plantains has been variously reported as low as zero6 and as high as 28.6X7 (Table 1). Most of that work was undertaken with undescribed cultivars of plantains. Studies are reported here in which the composition of fruits of three contrasting plantain cultivars and one banana cultivar were compared throughout ripening and senescence.
2. Experimental 2.1. Descriptionand supply of material

The plantains (Musa AAB group) used were all major Ghanaian cultivars from the collection of the University of Ghana, which have been described by Karikari.8-QApem was a typical cultivar of the French type with a low mean finger weight of about 200 g. The other two cultivars were both of the Horn type; Apantu had an intermediate mean finger weight, about 300 g, and Osa a high mean finger weight, about 400 g. Bananas (Musa AAA group) were purchased in the UK after commercial
0022-5142/81/1000-1021 $02.00

0 1981 Society of Chemical Industry

1022
Table 1. Composition of ripe plantains in different investigations

J. Marriott et a/.

Present work"

Stratton Sanchez Nieva and Loeseckels et al.6

Arriola Ketiku7
eta/.'*

Karikari
el

Fully ripe

Overripe

As percentage fresh pulp Total dry matter Total carbohydrates Starch Total sugars Non-reducing sugars Reducing sugars

36.2
30.6 11.7

38.9 21.5
0.1

43 . O 36 0 28.6 1.4
1.1

39.5
-

31.8

35.2 30.1
3.4

28 6
9.0

18 8

18.9
0.0

21.4 2.0 19.4

21.5 1.0 14.5

19.6

26.7 14.8 11.9

9.0 9.8

14.3 5.3

18.9

6.3

As percentage dry matter

85

55

84

__

16

86

Mean values for three cultivars.

shipment by sea from the Windward Islands; they were typical of fruits of the Cavendish subgroup, probably of the cv. Robusta. Plantains were harvested, packed and transported to the UK by air as described by Karikari et a l l 0 They were received 2-3 days after harvesting and were stored for not more than 7 days at 13 "C before ripening.

2.2. Ripening conditions Fruits were placed in 10-litre jars and were ripened at 20C. Initially, jars were sealed, ethylene injected to give a concentration of about 100 mg litre-1 and left for 24 h. Jars were then continuously ventilated with 10 litre h-l of humidified air and two fingers of each cultivar were removed at each sampling time. Fruits were sampled prior to ethylene treatment, on reaching colour stage 4 (more green than yellow) and wherever possible at 2-day intervals thereafter until they were very discoloured and senescent.

2.3. Analytical methods Colour score and pulp rupture force were measured with fresh fruit as described previously.11 Transverse slices were cut from the centre of both fruits, weighed, then frozen immediately on dishes placed in liquid nitrogen, lyophilised and reweighed to determine dry weight. The lyophilised material was extracted with 85% ethanol in a Soxhlet extractor and alcoholinsoluble solids were determined gravimetrically. Starch in the alcohol-insoluble material was hydrolysed with 0 . 7 HCl ~ on a boiling water-bath for 3.5 h12 and liberated glucose was determined using glucose oxidase (Boehringer Corp.).I3 Glucose, fructose and sucrose were determined in the ethanol solution, using test combinations for food analysis supplied by Boehringer Corp. For glucose and fructose, extracts adjust to pH 7.6 were phosphorylated with adenosine triphosphate in the presence of hexokinase followed by spectrophotometric determination of glucose-6-phosphate (G-6-P) by reaction with nicotinamide adonine dinucleotide phosphate in the presence of G-6-P dehydrogenase and then of fructose-6-phosphate by the same reaction after conversion to G-6-P with phosphoglucose isomerase. For sucrose, extracts were hydrolysed at pH 4.6 with 13-fructosidase and liberated glucose determined as described.13 Determination of individual sugars by high-

Starch and sugar transformation during ripening of fruits

1023

pressure liquid chromatography (h.p.1.c.) was carried out using a Partisil-10-PAC column with of sugars by paper chromatography was carried out xylose as an internal ~ t a n d a r d . 1 Separation ~ in butan-I-ol-acetic acid-water (5: 1 :4by VOI.).~

3 . Results
Changes in coloration, pulp softening and content of starch and individual sugars were similar in all three plantain cultivars; data for Apantu (Figures 1-3) were typical. Ripening of bananas and of plantains to colour stage 4 took at least 2 days but not more than 3 days after commencement of ethylene treatment. Thereafter, banana coloration was slower than for plantains but then became more rapid (Figure 1). Hereafter, fully ripe bananas and plantains refer to these fruits 4 days after colour stage 4 and overripe bananas and plantains to these fruits 8 and 12 days, respectively, after colour stage 4.

0 . 3-

Period a f t e r colour stage 4 (days)

(b)

Figure 2. Changes in alcohol-insoluble solids ( 0) and starch ( 0 ) during ripening of: (a) banana; and (b) plantain cv. Apantu. G indicates unripe, green fruits.

Rupture forces in the unripe fruits were 2.6, 2.35, 2.8 and 2.55 kg for bananas and cultivars Apem, Apantu and Osa, respectively, but once ripening commenced, plantains were always firmer than bananas (Figure 1). The dry matter (DM) content of bananas decreased from 31.8% when unripe to 29.3% when overripe and the mean DM content of the plantains decreased from 39.6% when unripe to 35.2% when overripe. This change is well documented and reflects changes in the osmotic balance between pulp and peel during r i ~ e n i n g Data .~ in Figures 2 and 3 are expressed as percentages of DM but data quoted in the text are expressed as percentages of fresh pulp for more ready comparison with other literature reports (Table I), which have commonly been expressed on this basis. Starch content (Figure 2) in fully ripe fruits was 0.8, 11.O, 8.3 and 7.3 % in banana and plantain cultivars Apem, Apantu, and Osa, respectively. In bananas, starch was very low (0.03 %) once they

1024

J. Marriott el rtl.

became overripe whereas the starch contents of overripe plantains were 4.0,3.6 and 3.8% for cultivars Apem, Apantu and Osa, respectively. Paper chromatography of alcoholic extracts of banana and plantain samples showed three major components corresponding to sucrose, fructose and glucose. When overloaded so that the main components did not resolve, a fourth component with a mobility intermediate between sucrose and raffinose was observed in bananas and each of the plantain cultivars. This component corresponded in mobility to a trisaccharide, detected in bananas by Henderson et al.5 and tentatively identified by them as a /3-fructosylsucrose. Analysis of four selected samples of each cultivar by h.p.1.c. showed that sucrose, glucose and fructose were the only components which could be detected (Figure 3) and quantitative data for all these samples were in good agreement with those obtained

Figure 3. Separation of sugars by h.p.1.c. in an extract from plantain cv. Apantu, colour stage 4. A, xylose added as internal . 5 : / . (3.9%); standard; B, fructose, 3.6% (4.0%); C, glucose, 3 D, sucrose, 27.9% (26.8 %). Values in parentheses are those obtained by enzymic analysis of this sample.

lOOr
Y 80c )

'
2

6040-

1 0
12
1 4
Period a f t e r colour stage 4 (days)

Figure 4. Changes in total sugars (0) sucrose ( 0) and reducing sugars (glucose+fructose) ( 0 ) during ripening of: (a) banana; and (b) plantain cv. Apantu. G indicates unripe, green fruits.

by enzymic methods. The oligosaccharide was not detected by h.p.l.c., the sensitivity of the method being such that it must have represented less than 1% of the total sugars. Total sugar content (Figure 4)was derived from the sum of glucose, fructose and sucrose contents, measured by enzymic methods. Total sugar contents of fully ripe fruits were 23.5, 17.4, 20.0 and 21.5% for bananas and plantains of cultivars Apem, Apantu and Osa, respectively. Total sugar contents of overripe fruits were 23.1, 25.7, 26.3 and 28.1 %,respectively, in the same cultivars. Thus, total sugar content did not change significantly during the later stage of ripening in bananas but it increased by about 7 % in plantains.

Starch and sugar transformation during ripening of fruits

1025

The ratio of g1ucose:fructose was between 0.98 and 1.05 for bananas and for plantains at all stages of ripeness except in the unripe fruit where concentrations were low. In both bananas and plantains, sucrose content reached a maximum early during ripening whereas that of glucose and fructose increased continuously (Figure 4). Expressed as a percentage of total sugars, sucrose comprised 73, 75, 72 and 7 2 % in fully ripe fruits and 30, 45, 43 and 37 % in very overripe fruits for banana and plantain cultivars Apem, Apantu and Osa, respectively.
4. Discussion

The starch and sugar composition of ripe plantains varied widely within each cultivar according to the precise stage of ripening or senescence of the fruits, though the trend of changes was similar for each cultivar. It was apparent that the production of sugars was much slower in plantains than in bananas and that whereas the sugar content of fully ripe bananas was almost constant until the fruits were very senescent, the sugar content of plantains increased continuously. Precise control of the stage of ripening prior to utilisation is therefore a critical factor in processing of plantains. Ketiku7 reported the starch content of ripe plantains as 28.6% of fresh weight but sampled at only one time during ripening. Comparison with other investigators (Table 1) indicates that these fruits were probably only partially ripe and that further starch hydrolysis has been consistently observed when ripening is prolonged. The data on starch and sugar transformations obtained here are similar to those of Stratton and Loesecke,ls cited by L ~ e s e c k ewhere ,~ starch was determined by acid hydrolysis12 in both investigations, but differ from those of Sanchez-Nieva et aI.,6 where starch was determined by a colorimetric method based on the starch-iodine reaction l6 (Table 1). Acid hydrolysis of starch is a standard method, which should not be subject to large errors, though it may slightly over-estimate starch content if P-glucans are present.I7 The discrepancies between the observations of Sanchez-Nieva et aL6 and those of other investigators probably reflect errors associated with their method of starch analysis, which has not been calibrated against any standard hydrolytic method using banana or plantain starch. Whatever the source of these discrepancies, it is clear that with the ripening conditions used here, starch and/or other endogenously hydrolysed polyglucans persisted into senescence, in all three plantain cultivars, whereas in bananas, hydrolysis was virtually complete in fully ripe fruits. The proportions of sucrose and reducing sugars (glucose and fructose) were similar in bananas and plantains, whereas most authors6, 7 , l5 have reported non-reducing sugars as very low in plantains (Table 1). Here, specific enzymic methods have been used for determination of individual sugars, supplemented with h.p.1.c. determinations. One other investigator'* used h.p.1.c. to determine individual sugars and found levels of non-reducing sugars which approached those reported here, whereas other investigators have used reducing sugar methods coupled to acid hydrolysis of sucrose. However, the most likely source of these discrepancies is enzymic conversion of sucrose to glucose and fructose after sampling. In the present work, rapid freezing, freeze-drying and alcoholextraction was used to prevent this. Acknowledgements This work was carried out as part of a collaborative programme between the Department of Crop Science, University of Ghana and Tropical Products Institute and was financed in part by a TPI extra-mural contract to the University. The authors thank the British Food Manufacturing Industries Research Association for the use of h.p.1.c. equipment and for assistance in its operation. References
1.

2.
3. 4. 5.

Simmonds, N. W. Bananas Longmans, London, 2nd edn, 1966. Marriott, J. Bananas-Physiology and biochemistry of storage and ripening for optimum quality. Crit. Rev. Niitr. Food Sci. 1980, 13, 41-88. Anon. Bnnana Riperting Manrial United Fruit Sales Corp., Boston, Mass., 1964. Von Loesecke, H. Bunonas Interscience, New York 1949. Henderson, R. W.; Morton, R. K.; Rawlinson, W. A. Oligosaccharide synthesis in the banana and its relationship to the transferase activity of invertase. Biochem. J . 1959, 72, 340-344.

1026

J. Marriott ei ul.
Sanchez-Nieva, F.; Hernandez, I.; Bueso de Vinas, C. Studies on the ripening of plantains under controlled conditions. J . Agric. Univ. P.R. 1970, 54, 517-529. Ketiku, A. 0. Chemical composition of unripe (green) and ripe plantain (Musa parudisiaca). J . Sci. Food Agric. 1973, 24, 703-707. Karikari, S. K. A note o n plantain (Musa AAB Group) and banana (Musa ABB Group) cultivars i n Ghana. Ghana J. Agric. Sci. 1971, 4, 79-85. Karikari, S. K. Some taxonomic assessments of the contributions of Masa Acuminata and Musa Balbisiana to the origins of plantains and bananas in Ghana. GhanaJ. Agric. Sci. 1973, 6, 9-19. Karikari, S. K.; Marriott, J.; Hutchins, P. Changes during the respiratory climacteric in ripening plantain fruits. Sci. Hortic. 1979, 10, 369-386. New, S .; Marriott, J. Post-harvest physiology of Tetraploid banana fruit. Ann. Appl. Biol. 1974, 78, 193-204. Association of Official Analytical Chemists Oficial Methods of Analysis for the Association of Oficial Analytical Chemists (Horwitz, W., Ed.), Association of Official Analytical Chemists, Washington, D.C., 1965, 10th edn. Bergmeyer, H. U. Methods of Enzymatic Analysis Verlag Chernie and Academic Press, 1965, 2nd edn. Rabel, F. M. ; Caputo, A. G.; Butts, E. T. Separation of carbohydrates on a new polar-bonded phase material. J. Chromatogr. 1976, 126, 731-740. Stratton, F. C.; Loesecke, H. von A chemical study of different varieties of bananas during ripening. Bull, Res. Dept. United Fruit Co 1930, No. 32. Carter, G. H.; Neubert, A. M. Plant starch analysis. Rapid determination of starch in apples. J. Agric. Food Chem. 1954, 2, 1070-1072. Southgate, D. A. T. Determination of Food Carbohydrates Applied Science Publishers, London, 1976, p. 50. de Arriola, M.; Calzada, J. F.; Menchu, J. F. Some physico-chemical changes during the storage and ripening of plantains. froc. ACORBAT Con5 Panama, 1979, pp. 251-263.

6. 7. 8. 9. 10.

I I.
12. 13. 14. 15. 16. 17. 18.