PHYSICOCHEMICAL CHARACTERISTICS, POPULARITY AND SUSCEPTIBILITY TO ANTHRACNOSE OF SOME LOCAL BANANA CULTIVARS

The physicochemical characteristics, peel thiclcness, firmness, peel to pulp ralio, pI-I, % titratable acidity (%TA), and solublc solids content (SSC), werc determined in six local cultivars of banana; 'Ambon', 'Ernbul', 'Kolilruttu', 'Seeniltehel', 'Puwalu', and 'Anamalu'. Physicochemical characteristics were comparable to those of some commercially important cultivars in other countries. A tendancy for a positive correlation (pS0.1) was observed between pairs of physical parameters (peel thickness, firmness ancl pee1:pulp). The lowest and the highest peel thiclrnesses were observed in 'Kolilcuttu' (1.27 mm) and 'Anamalu' (2.73 mm) respectively. These cultivars also recorded the lowest ('Koli.lruttu'-0.91 kgcm'2) and tha highest ('Anamalu'-2.01 kg cm.3 firmness values. The highest peel to pulp ratio was in 'Anamalu' (0.24) and the lowest, in 'Kolilcuttu' (0.09). A significantly lower pII (3.73) ancl higher %TA (10.56) were observed in 'Emlnul'. The pulp of ripe 'Seenikehel' had a significantly higher SSC (25.58Rrix). A taste preference stucly showed that 'Kolilruttu' was the most popular, while 'Ernbul' and 'Ambon' werc preferred second, equally. The least prefkrred were 'Anamalu', 'Seenilcehel' and 'Puwalu'. Initiation of anthracnose was delayed the most, in'Pi~walu', when conidia of C. n.itLsac., from typical anthracnose lesions from 'Anamalu' f rui ts were inoculated onto ripe fruits of the other 5 cultivars. In 'Puwalu', anthracnose development and spreading of lesions were also comparatively slow. Anlhracnose development was most rapid in 'Kolilcuttu', which also recorded the lowest firmncss and the lowest peel to pulp ratio. The correlation between lesion diameters ancl each physical parameter was negative without statistical significance.


INTRODUCTION 1-2
About 22 cultivars of dessert bananas are available in Sri Lanka. Some of these may have a good export potential, either fresh or processed. Most of the commercial banana cultivars grown in Sri Lanka are susceptible to anthracnose causing considerable postharvest 10sses.''~ Globally, the banana trade has mainly centred around the 'Cavendish' type. This though available freely, catering to the tastes of different ethnic groups in industrialized countries has become lucrative business and it is time that a larger choice of banana varieties is introduced to the world.
Having a large collection of locally grown cultivars, Sri Lanka has t h e potential to offer a variety of new cultivars of banana. Sri Lankan banana cultivars may be comparable to the cultivars grown in the Asian region. There are no records available on physicochemical characteristics and susceptibility to anthracnose by these local cultivars.
This study was a n initial attempt to determine some postharvcst parameters of six locally popular banana cultivars with the intention of evaluating their potential popularity in the international market. Among the parameters determined were pl~ysicochemical attributes, their degree of popularity among Sri Lankans, and the degree of susceptibility to anthracnose (Colletotrzclzum mrrsae) by five most popular cultivars of banana as a measure of their keeping quality. The preliminary investigations on physicochemical parametcrs were presented earlier6, which helped us decide on parameters to be investigated in the present study. 111 addition correlation coefficients were calculated between pairs of physical parameters investigated, and between susceptibility to anthracnose and each of the physical parameters.

Bananas:
Six cultivars of dessert bananas grown in Sri Lanka, namely 'Ambon' (AAA), 'Embul' (AAB), 'Kolikuttu' (AAB), 'Seenikehel' (ABB) 'Puwalu' (AAB) and 'Anamalu' (AAA) were used in this study. All bananas were subject to the experiments described below, within 24 h of turning to fu1l.y yellow stage. In the case of'Anamalu' which does not turn yellow with ripening, changing into a lighter green skin colour was used as an estimate for timing of analysis.

Physicochemical characters:
Frrtit firmness: Three 1 cm thick transverse sections were removed from the pulp of each finger and firmness was recorded by using a hand held penetrometer (Forestry Suppliers Inc.,UK) as de~cribed.~

Peel tlzickness:
Bananas were peeled off and thickness of the peel was measured a t three different locations (stylar-end, middle a n d stem-end) by using a vernier caliper a n d t h e average value was determined.
Peel to prdp ratio: Three 1 cm thick transverse sections were removed from the centre of each banana and the diameter of'the entire section (i.e. including peel) was taken as d, and the diameter of the pulp was taken as d,. The peel to pulp ratio was calculated as (d, -d2) + d2.

pH:
A log sample of thc pulp was homogenized in a blender for 2 min. Approximately lml portion of this was rcmoved to dctermine soluble solids contcnt (SSC). Thc homogenized pulp was diluted 1:4 ( l o g i n 40 ml water). Thc pH of pulp was determined by inserting the clectrode of a pH metcr (TOA Electronic Ltd; Japan HM 205) to the dlluted pulp.7 O/o Titl-ntnhle acidity (%TA): Samples (10 ml) of the diluted pulp of which the pH was determincd, were titratcd against 0. I N NaOH (in triplicate) i n t h e presence of phenolphthalcin a s an indicator. The end point was taken as a sudden slight change to pink colour. The C/o TA was calculated by thc following equation : ' % ntratable acidity = volume of NaOH x N x 100 weight equivalant in grams.

S o l~~b l e Solids Contc?lzt (SSC):
The homogenate (1 ml) of pulp, apportioned before determining pH, was centrifuged a t 3000g for 5 min, and a drop of the clear supernatant was used to measure SSC in degrees Brix"using a hand held refractometer (Leica Model 10430).

Taste preference of cz~ltivars:
Eighty Sri Lankans within the age group of 20 to 55, who were familiar with the taste of all 6 cultivars of bananas i.e. 'Ambon', 'Embul', 'Kolikuttu', 'Seenikehel', 'Puwalu' and 'Anamalu', were asked to record their ordcr of preference for these cultivars, on a scale of 1 to 6, 1 indicating thc most preferred and 6 indicating the least preferred. If they had equal preference for two or more cultivars of banana, they were asked to give the same rank to them.

Susceptibility to Anthracnose:
PI-eparation of Inoczslz~ln of Colletotrichr~m mzrsae: C. musae was isolated from over-ripe 'Anamalu' fruits having typical anthracnose lesions. Approximately 2x2 mm2 peels from the growing edge of the lesion were cut. They were surface sterilized in a 1% NaOCl solution for 3 min and transferred onto a sterile filter paper to remove excess N~o c~. ' These sections were then plated on 1 0 to freshly prepared plates of Cook's No. 2 agar medium. The colonies of Colletot~~ichzrm muscle that grew on these plates were sub cultured on to freshly prepared Cook's No.2 a g a r plates. These were incubated for 6 days a t room temperature (i.e. 28+2C).
For preparing conidial suspcnsions for inoculation to bananas, conidia were harvested by gently scraping offthe colonies on the agar plates with a sterile spatula, after adding 10 ml of sterilized H,O. The suspension was filtered through glass wool. This conidial suspension was adjusted to a concentration of approximately 10" conidia ml-I.

Inoculation of frz~its:
Five, 20 p1 drops of the suspension were placed along the longitudinal axis of each banana using a micropipette, in a single row. Amoisture chamber was prepared by lining plastic trays (40 x 30 x 7.5 cm) with moist filter papers and on top of which clean dry, glass petri dishes were arranged to rest the fruits. The fruits were placed on the petri dishes without touching the water, and each tray was covered with a heavy glass plate. The fruits were kept i n moist chambers for 24-48 h a t ambient conditions. Susccptibility to anthracnosc w a s measured in t e r m s of t i m e t a k c n for initial disease appearance (indicated by browning of t h e skin) on t h e spot of inoculation, and the daily increase in lesion diameter.

Statistical Analysis:
A randomized complete block experiment design was used in each of the above studies. All experiments (except taste preference study) were repeated 3 times, with a t least 5 bananas representing a single cultivar in each trial. The statistical analysis for all of the above tests were performed by using a statistical 11 analysis package. If the F value by analysis of variance was significant, a Duncan multiple range test was performed for mean separation. The level of significance for all tests was 0.05.
Correlation coefficients (r) were calculated between pairs of physical parametcrs (i.e. peel thickness, firmness a n d peel to pulp ratio) a n d also between each physical parameter and disease susceptibility (taking diameters of anthracnose lesions on day 8 after inoculation, on t h e 5 cultivars tested a s a measure) a n d their statistical significance w a s determjnedJ%ased on a probability level of 50.1.
A tendancy for a positive correlation ( a t a significance of plO.1) was observed among phys~cal parameters [i.e. peel thickness and firmness (0.73), peel thickness and peel to pulp ratio (0.81)' and pcel to pulp ratio and firmness (0.83)]. 'Anamalu' had thc highest pccl thickness and firmness, and 'Kolikuttu', the lowest (Tablc 1). Thc highest and lowest pccl to pulp ratios were also recorded by 'Anamalu' and 'Kolikuttu' respectively. The pH values of the ripe pulp of'Ambon', 'Kolikuttu', 'Seenikehel', 'Puwalu' and 'Anamalu' were not significantly different from each othcr, but that of 'Embul' was 3.7 and this value was significantly low compared to the other five cultivars ( Table 2). The highest %TA of ripe pulp was also shown by 'Embul' (10.56%). The highest SSC values were from 'Seenikehel' (25.58" Brix) and 'Puwalu' (23.74 Brix) and the lowest was from 'Ambon' (19.67" Brix) ( Table 2). There was no significant difference between the SSC values of 'Anamalu' (21.78" Brix) and 'Kolilruttu' (2 1.86" Brix), and these values varied highly i n both these cultivars, indicated by 3 superscript characters following the SSC values (Table 2). The highest taste preference was for 'Kolikuttu' (which scored the lowest rank). 'Embul' and 'Ambon' were preferred next, with no significant difference between preferences for them. 'Seenikehel', 'Puwalu' and 'Anamalu' were the least preferred with no significant difference between their preferences (Tab1.e 3).  When inoculated with a culture of C. musae, initiation of disease was delayed the most, in cultivar 'Puwalu' where anthracnose lesions appeared after the 611' day of inoculation (Fig. 1). Anthracnose development, determined by spreading of the lesion size, was most rapid i n 'Kolikuttu'. When lesion sizes a t 8 days after inoculatioil were compared, therc was no significant difference (p I 0.05) in those of 'Ambon', 'Embul' and 'Seenikehel'. 'Puwalu' recorded the lowest, a~d -'Kolikuttu', the highest, valucs being significant a t 0.05 level. In 'Kolikuttu', several anthracnose lesions which devcloped a s n a t u r a l infections merged with the lesions being measured. Therefore lesion diameters in 'Kolikuttu' developing from the inoculated spots could not be measured after 8 days (Fig 1).
The correlation between lesion diameters and each physical parameter was negative (lesion diameter and peel thickness (-0.67), lesion diameter and firmness (-0.49), and lesion diameter and peel to pulp ratio (-0.59). None of these were statistically significant (p10.1) ( Table 4).

Tablc 4 : Correlation coefficients (r) among physical parameters and anthracnose lesion diameters, 8 days after inoculation.
Pecl thickness (rnrn) Firmness ( It was also noted that in 'Kolikuttu', loss of quality occurrcd due to finger drop and longitudinal splitting of pccl, thc fbrmer being morc frequent than the lattcr. Approxlmatcly 44% of t h c bananas used in the prescnt study exhib~ted finger drop within 72 h of turning fully ycllow. Both thcse phcaomcna always occurred after the fruits turncd fully ycllow.

DISCUSSION
I t is rccorded that the relative firmness of the banana fruit is greatly determined by physical and chemical attributes such as peel thickness and starch content." In the present study too a positive correlation between peel thickness and firmness (r=0.73, pC0.1) was observed.
Pecl thickness and peel to pulp ratio, as wcll as firmness and pcel to pulp ratio indicated a tendancy for positive correlation wlth pS0.1 (Tablc 4). The highcr pecl to pulp ratio and the thicker peel probably contributed to the highcr firmness of Anamalu (Tablc 1).
Thc highest and lowest sugar levels (inferred by the SSC) were recorded by 'Seenikehel' and 'Ambon' respectively. It was observed t h a t the SSC value of 'Kolilruttu' did not differ significantly from that of 'Ambon', as the variation of SSC in 'Kolikuttu' was higher ( Table 2). The PI-I value of tllc ripe pulp of 'Embul' was significantly low (3.73) while this cultivar had t l~e highest %TA (10.56) ( Table 2).
Although, the physicochemical characteristics (given in table 2 ) are reported t o contribute to tasteL" major contribution to flavour would be by volatile components. I t was beyond t h e scope of' t h i s s t u d y t o determine volatile components of the different cultivars. However, the results of the taste preference study was expected to compensate to a certain degree for this shortcon~ing.
'Anamalu' which h a d a low preference in t h i s study is recorded to be indigenous to S r i Lanka and shows resemblances to 'Gros ~i c h c l ' . '~l t h o u~h small quantities have been exported intermittently in the past, Gros Micliel 1s not 15 rccordcd as a commercially important cultivar. Besides, the green pecl colour when ripe, oS'Anamalu' may not be very appealing to the discrimiriatlng, affluent buyer, who would pay more for good quality produce. Moreover, 'Anamalu' did not excel in any of the physicochemical characteristics tested ( Table 2) which may partially contr~bute to taste, and its pecl to pulp ratio was the highest which ind~cates that there is a higher percentagc of the unusable peel.
'Anamalu' was not included in dctcrmining susceptibility to anthracnose as the C. InzLsae isolates for this study were obtained from authracrlose lesions ol' 'Anamalu' fruits. 'Anamalu' was chosen for thls purpose, as ~t rallked lowest in the taste preference study ( Table 3) and is a n unlikely caildldate to compete In tlic international market for reasons given above.
Although none of the correlation coefficients between lesion diameters and each ol'the physicochemical. parameters were significant, the strongest re1ationsh.i~ among them was between peel thickness and lesion diameters, with r= -0.67 (Table  4). 'Kolikuttu' recorded t h e largest lesion diameters (Fig 1) a n d h a d t h e least thiclcness (and firmness) of peel (Table 1). 'Puwalu' recorded the smallest; lesion diameters and h a d t h e thickest peel (cxclud.ing 'Anamalu' which was not tested for disease). Of both 'Puwalu' a n d 'Seenikehel', t h e l a t t e r h a d larger I esion diameters and significantly thinner peel, although, both had identical peel to pulp ratios (Table 1). 'Embul' 011, t h e other 'band, was sign-ificaatly less susceptible to anthracnose although it recorded the lowest peel tl~iclmess value which was not significantly different from that of 'Seenikehel' (Fig 1).
If the three-dimensional spreading of the disease was noted, instead ofthe lesion diameters, i t is possible that the negative trend between peel thickness and disease observed In this study would not exist. Yet, having a thicker peel may be advantageous as there is a delay in thc disease being visible, and its spreading to tl1e pulp would also be delayed. 'Ambon' which ranked second to 'Kolikuttu' in our taste preference study is reported to be the most important cultivar in Indonesia and is said to rank highly in ~a 1 a~s i a . l~ Considerable similarities were observed, when %TA and SSC values in this study were compared with different cultivars of commercially popular banana in the South East Asian countries. For instance, %TAvalue for the ripe pulp of 'Puwalu' (5.93%) compared with that of cultivar 'Pisang Kcpok' (6.14%) which is recorded as the most important cultivar in Indonesia and it was similar to the cultivar 'Saba' of Philippines, 'Pisang Nipah' of Malaysia and 'Kluai Hin' of Thailand". Similarly I SSC value obtained for rlpe pulp of 'Ambon' (19.67) was close to the value of the 18 cultivar 'Latundan'(20.0) . However, 'Latundan' has also becn compared with '~o l i l r u t t u ' . '~ The SSC value fbr ripe pulp of 'Secnikcl~cl' (25.6), which was the highest among the 6 cultivars, was comparable to that of cultivar 'Pisang Kcpok' (23.9) of Indonesia, which is recorded to have a high level of SSC.'" 'Embul' appears presently to be the best suited to the export market than 'Kolikuttu'. Its higher resistance to anthracnose is a n added advantage to preference. I t is rccorded to have a high export potential in the European 20.7 market.
The potential of less popular bananas to be processed or marlreted as vegetables need further investigation. Equivalents of 'Seenikehel' in other countries are recorded to be most widely grown and common i n those countries and in Thailand it is recorded to bc eaten frcsh or cooked.l7 Thcrefore ~t may be possible to popularize this cultivar as a vegetable than a fruit. Furthermore, its high sugar content may makc it suitable as dricd ingredic~lts in desserts, snacks, etc., or processed as flakes or purees.
The results obtained from this study show that some local cultivars may compete well with commercially important bananas. Further studies are necessary to improve keeping quality of locally popular bananas, such as 'Kolikuttu', with emphasis on developing resistance to anthracnose. Although this was the most popular cultivar, it was very susceptible to anthracnose. In addition, fully ripe fruits were susceptible to finger drop and less frequently to longitudinal skin splitting.