Ice cream and frozen desserts are mainly valued for their pleasing flavor, cooling effects and refreshing tastes. The awareness of consumers for healthier and functional food has led to the introduction in ice cream manufacture of certain materials with documented nutritional and physiological properties such as probiotics [1,2], lactic acid bacteria [3], dietary fibers [4], alternative sweeteners [5], natural antioxidants [6] and low glycemic index sweeteners [7]. In today’s food industry, a global trend towards the manufacture of healthier and more natural fruit and vegetable food products, such as soups, smoothies and sauces, is ongoing, as well as the incorporation of puréed vegetables in other food products. Currently health is a major concern of customers. Therefore, manufacturers are finding new ways to incorporate natural and innovative ingredients into ice cream for health benefits.

Vegetable plays an important role in daily human diet. Consumption of vegetables has been associated with protection against certain types of cancer, cardiovascular disease and various age related diseases [8]. «Thy food be thy medicine» is one of the laws of nature. WHO and FAO launched [9], a joint initiative to promote fruits and vegetables for health worldwide [10]. The role of vegetables in the human diet has increased since they provide essential carbohydrates, proteins, fiber, vitamins and minerals. They are important sources of vitamins, minerals and salts required for human nutrition. The carbohydrates, proteins and fats are required comparatively in larger quantity than vitamins and minerals. They also supply carbohydrates for energy and proteinous compounds for muscles building. They also contribute to roughage and fiber. Vegetables not only form an essential part of a well-balanced diet, but the flavor, aroma, also make them important in human diet and appetite [11]. WHO [9] recommends a minimum of 400 g of fruits and vegetables per day (excluding potatoes and other starchy tubers) for the prevention of chronic diseases such as heart diseases, cancer, diabetes and obesity as well as for prevention and alleviation of several micronutrient deficiencies, especially in less developed countries. In India more then 20-25% of fruits and vegetables are spoiled before utilization. Despite being the world’s second largest producer of fruits and vegetables, only 1.5% of the total fruits and vegetables produced are processed. Incorporation of vegetable for the preparation of dairy products is thought to be a convenient and economical alternative for utilization of these vegetables [12]. Hence, there is a need to develop new functional dairy products to reflect consumer interest in health (e.g. utilizing vegetable source with phytochemicals) and naturalness [13]. Among various vegetables grown in India, bottle gourd (Langenaria siceraria) has a high place in diet as it is rich and the cheaper source of nutrients. Bottle gourd has long been an important component of indigenous herbal medicine, particularly in Asia [14]. Bottle gourd was used as an ingredient in ice-cream in this study since it is a highly valued vegetable containing good amounts of nutrients like carbohydrates, Vitamin A, Vitamin C and minerals.

Whole milk (5.5 ± 0.13% fat; 8.5 ± 0.16% MSNF; titratable acidity 0.16 ± 0.02% LA) procured from Vidya Dairy, Anand and “Sagar” brand skim milk powder (SMP), marketed by Gujarat Cooperative Milk Marketing Federation Ltd., Anand were used as the base materials for ice cream manufacture. Commercial grade cane sugar (M grade) was obtained from the local market of Anand. Sodium alginate (RM- 7494), guar gum (RM-1233) and carrageenan (RM-1576) from Hi media Laboratories Pvt. Ltd., Mumbai were used as a stabilizers and commercial grade Glycerol Mono Stearate (GMS) (Brion Fine Chem., Mumbai) was used as emulsifier. Vanilla essence No.1 (M/s Bush Boake Allen (India) Ltd., Chennai), cardamom powder (freshly made from commercially available cardamom) and saffron essence (Bush International Flavors and Fragrances India Pvt. Ltd., Mumbai) were used as the flavoring agents in bottle gourd ice cream. Bottle gourd (Pusa Naveen and Pusa Summer variety) were procured from local vegetable market of Anand and Anand Bottle Gourd – 1 variety was procured from the Anand Agricultural University farm located at Anand, Gujarat.

Selection of variety of bottle gourd

Three varieties of bottle gourd viz. Pusa Summer, Pusa Naveen and Anand Bottle gourd-1 were screened for their suitability for preparation of bottle gourd ice-cream. Processed bottle gourd cubes and shreds were prepared using the procedure described by Babar [15] and processed bottle gourd puree was prepared using the procedure described by Agarwal and Prasad [16]. Processing of bottle gourd with peel resulted in bitter taste and off flavor in the final processed bottle gourd form than prepared from without peel bottle gourd. Therefore, peeled bottle gourd was used in the bottle gourd preparations in this study. The products prepared using three varieties were subjected to ranking test for preference using a panel of 25 judges selected from employees and students within department. The summary of ranking results of preference of processed cubes, shreds and puree prepared using three varieties of bottle gourd is presented in Table 1.

Table 1: Summary of ranking results of preference of processed cubes, shreds and puree prepared using three varieties of bottle gourd.

There was a significant difference (P<0.01) in preferences between the products obtained from the three varieties of bottle gourd studied. Products prepared using Pusa Naveen and Anand Bottle gourd-1 were significantly (P<0.05) preferred compared to products prepared using Pusa Summer variety. The products prepared using Pusa Naveen (and/ or Anand Bottle gourd-1) were dark green in color, optimum chewiness and firmness and pleasing clean caramelized bottle gourd flavor. On the other hand, products prepared using Pusa Summer were light green in color, hard body, more chewy texture and lacked optimum bottle gourd flavor and were sometimes criticized for being slightly bitter in taste. It can be seen from Table 1 that there was no significant (P>0.05) difference in preference of products viz. cubes, shreds and puree prepared from Pusa Naveen and Anand Bottle gourd-1. However, since Pusa Naveen variety is available in the market throughout the year, it was selected and used throughout the study. Firm, tender, uniform, green, nearly straight and cylindrical bottle gourds, at a commercially marketable stage were obtained from a local market in Anand, Gujarat. Bottle gourds having 30 to 40 cm length and 8 to 10 cm thick were selected. Bottle gourds of cylindrical shape and having light green to slight dark greenish but uniform in color were selected. The average weight of selected bottle gourds ranged from 0.5 kg to 1 kg.

Preparation of sugar treated bottle gourd shreds

Bottle gourds were peeled, washed and halved into two parts using clean, dry knife, de-cored and shredded (1.6 ± 0.1 cm length and 0.3 ± 0.1 cm width and thickness) using a shredder attachment (Model-Boss Food Processor, Boss Electrical, Solan and Himachal Pradesh). From 1 kg bottle gourd ~ 15 to 20% of its weight was peel and 20 to 25% was core portion. Therefore, quantity of edible portion bottle gourd obtained after peeling and decoring was 40 to 45% of weight of initial weight of bottle gourd. The shreds were blanched (90oC for 30 s) to prevent enzymatic browning [17] and transferred in pre-washed, sanitized clean pre-weighed stainless steel vessel. After draining, the shreds were mixed with water (obtained during blanching) and sugar at the rate of 70% by weight of raw bottle gourd shreds respectively. Sodium chloride was added at the rate 0.1% by weight of raw bottle gourd shreds and cooked on a medium flame till concentration of sugar syrup reached 60, 65, 70 and 75o Brix. In preliminary studies, it was found that concentration up to 60 and 65o Brix resulted in final shreds having uncooked, predominant raw bottle gourd flavor and less acceptable in terms of sensory characteristic was observed. Concentration to 70o Brix resulted in final quality shreds having very pleasant, clean, caramelized, properly cooked, no raw bottle gourd flavor, good texture was seen but above 70o Brix resulted in excessive increase in caramelized flavor, browning and chances of sticking or burning was more due to excessive concentration. Therefore, 70o Brix was used in preparation of shreds. To prepare 1 kg processed bottle gourd shreds, approximately 1 kg peeled and decored bottle gourd was required.

Preparation of sugar treated bottle gourd cubes

Bottle gourds were peeled, washed, halved into two parts, de-cored and cut in pieces (average length 5.0 ± 1.0 cm slices, breath 2.0 ± 0.5 cm and thickness 2.0 ± 0.5 cm). These raw pieces were blanched (90oC for 2 min) and drained. The pieces were then cut into 0.8 ± 0.2 cm3 cubes and taken in pre-washed, sanitized, clean, pre-weighed stainless steel vessel. Water and sugar (70% by weight of bottle gourd cubes respectively) and salt (0.1% by weight of cubes) were added to the bottle gourd cubes and concentration was done till different level of degree Brix of sugar syrup (using digital Refractometer) was achieved. Different sugar syrup concentration was studied viz. 60, 65, 70 and 75o Brix by heating on medium flame. It was observed that cooking to final concentration up to 60 and 65o Brix resulted in final cubes having uncooked, unpleasant, predominant raw bottle gourd flavor, less sweet and less acceptable in terms of sensory characteristics. Cooking to concentration up to 70o Brix resulted in final quality cubes having very pleasant, clean, caramelized, sweet flavor and properly cooked firm texture was seen but above 75o Brix resulted in excessive increase in highly cooked, caramelized flavor, browning and chances of sticking or burning was more due to excessive concentration. So, 70o Brix was used in the final preparation of cubes [15]. Hence, 70o Brix sugar syrup concentration was used in the final preparation of cubes. To prepare 1 kg processed bottle gourd cubes 1 kg peeled and decored bottle gourd was required.

It has been reported in literature that treating bottle gourd cubes with calcium chloride solution for preparation of tutti fruity from bottle gourd resulted in improved firmness of cubes [15]. Therefore in this part, effect of treating blanched cubes by addition of calcium chloride on body and texture of processed bottle gourd cubes were evaluated. Control was also prepared without CaCl₂ treatment. After blanching the cubes were dipped in 1% calcium chloride solution for 3 h. The cubes were then drained, mixed with equal quantity of 50o Brix sugar syrup and cooked by heating on a direct flame till the concentration of sugar syrup reached 70o Brix. It was observed that CaCl₂ treated cubes had slight off flavor. However, the shape and color were acceptable. The control sample had good shape, flavor and color retention, uniform texture and firmness like tutti fruity. No perceptible difference was found in body and firmness of the samples. CaCl₂ treatment caused no improvement on texture or firmness of bottle gourd cubes. Hence, CaCl₂ treatment was not used during preparation of bottle gourd cubes.

Preparation of processed bottle gourd puree

Bottle gourds were peeled, washed and cut into pieces of average length 5.0 ± 1.0 cm slices, breath 2.0 ± 0.5 cm and thickness 2.0 ± 0.5 cm using a clean sharp, dry stainless steel knife. These pieces were immediately blanched (90oC for 2 min) to prevent enzymatic browning which affects the final quality of puree in terms of color and other sensory aspects [17]. Puree was made from the blanched and drained bottle gourd slices by grinding for 2 to 3 min food processor (Model- Boss Food Processor, Boss Electrical, Solan and Himachal Pradesh). The puree was then transferred in a clean, sanitized pre-weighed stainless steel vessel and sugar was added @ 50% (by weight of puree) and blended. In preliminary investigations it was found that addition of sugar at a higher rate resulted in excessive browning and pronounced caramelized flavor whereas addition of sugar at lower rate resulted in a product with decreased sweetness and unpleasant predominant bottle gourd flavor. The contents in the vessel were then cooked on medium flame. In order to select the concentration ratio (CR) with a view to achieve optimum flavor, color and acceptability of the final puree, three concentration ratios were employed viz. 1.3, 1.5 and 1.7. A properly cooked, clean, pleasing, caramelized bottle gourd flavor with optimum sweetness and dark green in color and having highly acceptable optimum flavor was obtained when a CR of 1.5 was used. Whereas use of lower CR i.e. 1.3 resulted in a puree with predominant raw bottle gourd flavor, lacking desired consistency, lighter in color, is lacking in sweetness. On the other hand, use of higher CR i.e. 1.7 resulted in highly caramelized flavor; excessively sweet, brown color and slight bitterness. Product became sticky therefore chances of browning and burning were more. Hence a CR of 1.5 was selected and used in preparation of bottle gourd puree. To prepare 1 kg processed bottle gourd puree 1.1 kg peeled bottle gourd was required. The prepared bottle gourd puree was bland in taste. Hence, sodium chloride was added at the rate of 0.1% by weight of raw puree to enhance its flavor.

Preparation of ice cream mix

Whole milk, cream and SMP used in the manufacture of ice cream mix were analyzed for their composition. The quantity of milk, cream, SMP, sucrose, sodium alginate, guar gum, carrageenan and GMS required for a batch (i.e. 5.0 kg of ice cream mix to be frozen in a direct expansion type batch freezer) was calculated using serum point method [18]. Cream (40.0 ± 2% milk fat separated from whole milk) was mixed with calculated quantity of whole milk and pre-heated to 45°C. All the dry ingredients viz. SMP, sugar, stabilizers and emulsifiers were dry blended and added to the milk - cream mixture at a temperature of about 50°C. The mix was further heated to 75°C and subjected to homogenization (150 and 50 kg/cm2 pressure in the first and second stage respectively, M/s. Pal Engineering Ltd., Ahmedabad). The mix was then pasteurized at 80°C for 5 min prior to cooling and aging overnight at 3 ± 1ºC. The flavoring ingredients were added just prior to freezing.

Preparation of ice cream

For preparing different batches of ice cream in direct expansion type batch freezer (cylinder capacity 10.0 lit), the aged mixes were frozen in a horizontal batch freezer (M/s. Pal Engineering Pvt. Ltd., Ahmedabad) having an arrangement for air incorporation under pressure. The temperature of the circulating refrigerant was -23.0 to -30.0°C. After freezing the mix to a semi-solid consistency (10.0-15.0 min), as inferred from the load on the ammeter (initial beater load 2.1 amp; final load 2.6 amp), air was whipped in the freezer barrel at a pressure of 10 to 15 psi for about 2 min. Bottle gourd puree was incorporated in ice cream mix prior to freezing while the bottle gourd cubes were incorporated into the partially frozen ice cream before incorporation of air under pressure. The ice cream at the right stage of freezing, as ascertained from the consistency and overrun, was drawn directly into 100 ml High Impact Polystyrene (HIPS) ice cream cups and covered with wax coated paper board lids. The temperature of the ice cream at the drawing stage was recorded. Mixes were frozen to about -4.5 to -5.0°C and the targeted overrun was set at 90%. The filled ice cream packs were then transferred immediately to a hardening tunnel maintained at -25 ± 2°C and hardened for 3 h.

The hardened ice creams were then subjected to compositional analysis, melting quality test and sensory evaluation.

Analysis

Physico-chemical analysis of ice cream and ice cream mixes: The fat content of the ice creams were determined by the standard method as suggested in ISI Hand Book [19] for ice cream mixes using 5 g ice cream mix. The total solids of the ice cream were determined by the standard procedure as described in Laboratory Manual [20] using about 2 g of sample. The titratable acidity of the ice cream was determined by the standard method suggested in ISI Hand Book [19]. The protein content of the ice cream mixes was determined by Semi – Micro kjeldahl method [21], using Kjel-plus Digestion System (Model-KPS 006L, M/s.Pelican Instruments, Chennai) and Kjel-plus Semi-Automatic Distillation System (Model-Distil M, M/s.Pelican Instruments, Chennai). As per this method the total nitrogen was determined and the value so obtained was multiplied by a standard factor 6.38 to get the protein content. Ash content of ice-cream and icecream mixes were determined by procedure described in IS: 1547-1985 [21]. The calcium content of frozen product samples was determined in accordance with the method of Kindstedt and Kosikowski [22]. Potassium and phosphorus concentration were estimated by atomic absorption spectrophotometry (Spectrophotometer-Perkin Elmer Model 3110, λ=766.5 and 214.9 nm respectively) by AOAC [23,24]. The β-carotene content of ice-cream mix was determined by the method described in IS: 15120-2002 [25] and AOAC-2001.13. The crude fibre content of ice-cream mix was determined by the method described in EC No. 152/2009 [26]. The viscosity of ice cream mix was determined by the method of Loweenstein and Haddad [27] using a Brookefield Viscometer, Model DV-II + Pro (Brookefield Engineering Laboratories, USA). The viscosity readings were taken at 4°C after ageing mixes at 3-4°C for about 24 h. The specific gravity of the ice cream and ice cream mixes was determined at 20°C using a specific gravity bottle according to the method described by Ling [28]. Overrun was determined according to the method given by Marshall [18]. A known volume of ice cream mix was weighted accurately and then the same volume of ice cream was weighed and the overrun was calculated. The melting characteristics of ice cream were evaluated according to the method given by Loweenstein and Haddad [27] with slight modifications. A day prior to melting characteristics determination, samples were transferred to a -15 ± 2°C deep freezer and left overnight. One lit packet of ice cream was taken and a slice weighing 100 g was cut in duplicate. The slices were separately placed over a wire mesh screen (250 pores per sq. inch) and then placed over a long stem glass funnel of 6 inches diameter. The funnel with the wire meshes containing the ice cream slices were placed over a 100 ml glass cylinder. It was then kept in an incubator maintained at 30°C for 45 min. After 45 min the weight of ice cream melted was noted. The melting characteristics were determined as% of total ice cream melted in 45 min at 30°C. The hardness of the hardened frozen product was measured using FPN3 cone penetrometer (Associated Instrument Manufacturers Pvt. Ltd., India). The mass of the dropping assembly (shaft+cone) was 155 g. The hardened samples were subjected to penetration measurements at -15 ± 2°C. The samples were tested for the hardness by adjusting the cone of penetrometer exactly above the surface of the sample and allowing the cone to freely penetrate the product for 10 s. The depth of penetration was measured in 0.1 mm units on the dial of the instrument. These penetrations were made on each sample at three different points and the closest two readings were averaged. Total viable count, yeast and mold count and coliform count of ice-cream was determined by following the method described by Messer [29], Frank [30] and Hartman [31] respectively.

Sensory analysis

The ice cream samples were tempered to -12 ± 2°C for 1-2 h before judging. Sensory evaluation of the ice cream samples was conducted in isolated booths illuminated with incandescent light and maintained at 23 ± 2°C. Samples were served in the 100 ml polystyrene cups in which they were frozen. The cups were labeled with three-digit codes. The order of presentation of the samples was randomized across subjects. Subjects judged a maximum of 4 samples in one session. The sensory panel (n=7) was composed of staff members and post graduate students working in the institution. The ice cream was subjected to sensory evaluation using 100 point scale based score card suggested by Arbuckle [32].

Statistical analysis

Response Surface Method (RSM) using Design Expert 9.0.2 software was used for the development of bottle gourd ice cream formulation and process standardization. For selecting compatible flavor/s in formulation of bottle gourd ice cream, statistical analysis of the data was carried out as per Steel and Torrie [33] using completely randomized design.

Selection of most suitable form of processed bottle gourd for bottle gourd ice cream

To select the most suitable forms of processed bottle gourd for preparing acceptable quality bottle gourd ice-cream, ice cream was prepared in a commercial batch freezer using the various prepared forms viz. puree (4, 6, 8%), shreds (2, 4, 6, 8%), cubes (2, 4, 6, 8%) and combination of the two forms of bottle gourd viz. shreds + puree (2+10, 4+8, 6+6, 8+4%) and cubes + puree (2+10, 4+8, 6+6, 8+4%). Ice cream was prepared according to procedure described above. The various forms of bottle gourd at selected levels were added after aging.

It was found that addition of cube as well as shreds alone resulted in a product which was not acceptable in sensory properties and was less acceptable because addition of shreds or cubes alone resulted in a product with crunchy texture, lacking in flavor and lacking in desired caramelized, pleasant flavor. Addition of puree alone was also not liked well by panelists as it gave unacceptable flavor and faster meltdown. The combination of shreds and puree also had poor acceptability as it gave a similar effect as in case of shreds alone but had a better appearance. Addition of cubes in combination with puree was best in terms of acceptance. It was found that puree contributed to viscosity, richness and cubes contribute good flavor, appearance, texture and increase in palatability of ice cream. Hence, a combination of cubes and puree was selected and used.

Formulation of basic ice-cream mix for manufacture of bottle gourd ice-cream

The tentative levels of milk fat as well as MSNF were based on FSSA [34] requirements for different types of ice creams viz. low-fat, mediumfat and ice cream. In preliminary trials, ice cream mixes were prepared with 2.0, 6.0 and 12% milk fat, which corresponded to requirements for low-fat, medium-fat and regular ice cream respectively. The MSNF content of ice cream varies inversely with the fat content [18]. Therefore, the levels of MSNF selected were 12.0, 11.5 and 11.0% respectively. These levels were based on the minimum requirements for milk protein in ice cream mix according to FSSA [34] requirements. Sugar, sodium alginate and GMS were added at the rate of 15, 0.2 and 0.2% respectively in all the mixes. Treated bottle gourd puree was blended in mix at 4% level just before freezing whereas treated cubes were added at 8% using fruit feeder after drawing of ice-cream from batch freezer.

Since, it has been reported that bottle gourd contains pectin [35] and the fiber content of peeled bottle gourd was 0.7 ± 0.01% and, it was hypothesized that addition of bottle gourd puree would result in increased viscosity of the mixes and could partly be used as a fat replacer in ice cream. However, it was observed that the low-fat and medium-fat ice cream had very poor acceptability as the ice cream lacked desired richness, consistency, mouth feel, viscosity and had faster melting rate, watery body, ice crystals which could be perceived in mouth. At levels of 12.0% milk fat and 11% MSNF the ice cream was having rich mouth feel, smooth texture, soft body, good meltdown and good color and appearance. Therefore, a milk fat level of 12.0% and 11.0% MSNF was chosen in the formulation.

To select the level of sucrose in the formulation in the preliminary screening sucrose was used at the rate of 13, 14 and 15% (w/w) in the formulation of the frozen product. This sugar which was used for preparation of bottle gourd puree was not deducted from the total sugar required for preparation ice cream mix batch with different levels of sugar viz. 13, 14 and 15%. The sugar used in preparation of cubes was not deducted from the total sugar required for preparation ice cream mix as it did not contribute to overall sweetness of mix since it didn’t diffuse out from shreds since the sweetness contributed by cubes was noticed only during chewing. It was found that addition of sugar at higher levels viz. 15% resulted in ice cream which was too sweet and lower level viz. 13% was lacked in optimum sweetness. Hence, a level of 14% sucrose was chosen in the formulation.

It was found that the products prepared using sodium alginate (0.20%) alone as stabilizer had lower melting resistance, weak body and coarse texture. Ice cream manufacturers use blend of several stabilizer ingredients to achieve the desired characteristics which cannot be provided by a single ingredient [36]. Therefore preliminary investigations were carried out using different blend of stabilizers viz. Sodium alginate, guar gum, carboxymethyl cellulose and carrageenan. Addition of blend of sodium alginate, guar gum and carrageenan (0.1:0.05: 0.02%) along with 0.2% GMS as emulsifier resulted in a product having acceptable flavor, body and texture, good meltdown and total score. From the experimentation carried out, the formulation chosen was 12% milk fat, 11% MSNF, 14% sucrose, 0.1% sodium alginate, 0.05% guar gum, 0.02% carrageenan and 0.2% GMS, 37.37% total solids.

Optimizing the level of processed bottle gourd forms viz. cubes and puree in bottle gourd ice cream using Response Surface Methodology

Ice cream was made using processed bottle gourd cubes and puree using a Response Surface Methodology (RSM) to optimize the level of processed bottle gourd forms in bottle gourd ice cream. Response surface methodology (RSM) is an effective tool for optimizing a variety of food processes [37-39]. The main advantage of RSM is reduced number of experimental runs that provide sufficient information for statistically valid results. The RSM equations describe effects of the test variables on the observed responses, determine test variables interrelationships, and represent the combined effect of all test variables in the observed responses, enabling the experimenter to make efficient exploration of the process.

Ice cream was prepared by incorporating processed bottle gourd cubes and purees @ 6 to 8% and 2 to 6% (w/w) of mix respectively. The bottle gourd ice-cream prepared with different levels of bottle gourd cube and puree addition as shown in design matrix (Table 2) of two factor CCRD design consisting of 13 experiments were evaluated for various physico-chemical viz. TS, fat, ash and sensory characteristics. Vanilla was used as background flavoring agent in bottle gourd icecream throughout RSM experiments. Vanilla essence was added in ice cream mix at the rate of 1.5 ml per kg i.e. around half the rate required for flavoring plain vanilla ice cream.

Table 2: Experimental Design Matrix (CCRD) for Levels of factors: Bottle gourd Cubes () (A) and Bottle gourd Puree () (B).

Effect of different levels of bottle gourd cubes and puree on sensory properties of bottle gourd ice-cream

The sensory parameters chosen to assess the quality of bottle gourd ice-cream are flavor, color and appearance, body and texture, melting quality and total score. The ice cream samples were tempered to -12 ± 2o C for 1 to 2 h before judging. The sensory data along with formulations as per their run order is shown in Table 3. The scores of flavor, body and texture, color and appearance were well fitted in quadratic model. The quadratic model for sensory parameters namely flavor, color and appearance, body and texture, melting quality and total score were obtained through successive regression analysis. The calculated F values more than the Table F values at 5% level of confidence and it indicated the significance of model terms. Furthermore, the coefficient of determination (R²) that reflects the proportion of variability in data explained or accounted by the model for flavor, color and appearance, body and texture, melting quality and total score were 0.8110, 0.9425, 0.8238, 0.8514 and 0.9182 respectively. A larger R2 values suggest a better fit of the quadratic model. The adequate precision measures the signal to noise ratio, the value of which should be greater than 4.0. The adequate precision value for flavor, color and appearance, body and texture, melting quality and total score were 6.289, 16.221, 8.290, 9.607 and 11.056 respectively, which were greater than 4, hence supporting the suitability of model to navigate the design.

Table 3: Experimental Design Matrix and Scores for Sensory Properties of Bottle gourd Ice-cream

The ice-cream prepared by using 7% cube and 4% puree addition was rated best for flavor (Table 3 and Figure 1) whereas 8% cube and 6% puree addition was rated lowest score for flavor by the panelists. The coefficient of determination (R2) is the proportion of variability in the data explained or accounted for the model and high value (0.811) of R2 indicate a better fit of the model to the data. Also the adequate precision value of 6.289 recommends the use of this response to navigate the design. The P value of% puree (P<0.01) and cube (P<0.05) addition showed its significance at quadratic levels (Table 4). Increase in rate of addition of cube and puree in combination decreases the flavor score significantly (P<0.01) at higher level while at increasing cube and puree levels, flavor score was enhanced at linear level but at higher(quadratic) level again fl avor score decreased. Th e Multiple Regression equation generated to predict the fl avor as aff ected by diff erent factors in terms of coded factors is as follows:

Figure 1: Response surface of flavour as influenced by level of per cent bottle gourd cube and puree.

Flavor = + 40.02 + 0.39^* A + 0.023^*B-1.01^* AB-1.27^* A² -1.46^* B²

Where, A and B refer to% cube and% puree respectively. A signifi cant negative eff ect on fl avor was found with the higher rate of cube (P<0.05) (A²) and puree (P<0.01) (B²) addition in bottle gourd ice-cream. Th e square of factor indicates that the eff ect of ingredient at highest (quadratic) level used in the product standardization. A nonsignifi cant positive eff ect on fl avor was observed with both rate of cube (A) and puree (B) at linear level. Th e negative interaction indicates that both the ingredients are negatively correlated and by increasing the level of one ingredient, the level of other ingredient has to be decreased. A non-signifi cant negative interactive eff ect on fl avor was observed with both rate of cube and puree addition (AB).

Color and appearance of the product is the visual perception, which in turn, defi nes product quality. Th e color and appearance score of bottle gourd ice-cream ranged from 4.074 to 4.664 and the ice-cream prepared by using rate of 7% cube and 1.1715% puree addition scored lowest color and appearance score (Table 3 and Figure 2). Ice-cream prepared by using rate of 7% cube and 6.828% puree addition score was rated best for color and appearance by the panelists. Th e P value of% puree addition showed its signifi cance at linear and quadratic levels (Table 4). Ice-cream prepared with higher levels of cube and puree addition content were having dull or non-uniform color and appearance was not liked by panelists. Th e Multiple Regression equation generated to predict the color and appearance as aff ected by diff erent two factors in terms of coded factors are as follows:

Figure 2: Response surface of colour and appearance as influenced by level of per cent bottle gourd cube and puree.

Table 4: P Values and Partial Coeffi cients of Regression Equation of suggested models for sensory properties of Bottle gourd Ice-cream .

Color and Appearance = + 4.51- 0.042A + 0.18B- 0.037AB-8.606E - 003A² - 0.074B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.01) positive eff ect on color and appearance was observed with rate of puree (B) addition in ice-cream at linear level but signifi cant (P<0.01) negative eff ect was observed at quadratic level (B²). A non-signifi cant negative eff ect on color and appearance was observed with rate of cube at linear level (A) and quadratic level (A²). Th e negative interaction indicates that both the ingredients are negatively correlated and by increasing the level of one ingredient, the level of other ingredient has to be decreased. A non-signifi cant negative interactive eff ect on color and appearance was observed with rate of cube and puree addition (AB) in ice-cream.

Body is the standing properties of food. Texture is that property of food which is associated with the sense of feel or touch experienced by fi ngers or the mouth. Bottle gourd ice-cream prepared by using 8% cube and 6% puree addition was rated best for body and texture by the panelists (Figure 3). Th e P value of rate of puree addition showed its signifi cance (P<0.01) at linear and quadratic levels (Table 4). Th e Multiple Regression equation generated to predict the body and texture as aff ected by diff erent factors in terms of coded factors is as follows:

Figure 3: Response surface of body and texture as influenced by level of per cent bottle gourd cube and puree.

Body and Texture = + 26.70 + 0.29A + 0.66B- 0.34AB+ 0.037A² - 0.68B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.01) positive eff ect on body and texture was observed with rate of puree at linear level (B) while signifi cant (P<0.01) negative eff ect on body and texture was observed at quadratic level (B² ). A nonsignifi cant positive eff ect on body and texture was observed with rate of cube addition in ice-cream at linear level (A) and quadratic level (A²). Th e negative interaction indicates that both the ingredients are negatively correlated and by increasing the level of one ingredient, the level of other ingredient has to be decreased. A non-signifi cant negative interactive eff ect on body and texture was observed with rate of cube and puree addition (AB) in ice-cream.

Meltdown behavior provides insights into key ice cream phenomena, such as protein stability, fat agglomeration and air cell size, which infl uence body, texture, and other sensory characteristics. Bottle gourd ice-cream prepared by using 7% cube and 1.171% puree addition was rated best for melting quality by the panelists (Figure 4). Th e P value of rate of puree and cube addition showed its signifi cance (P<0.05) at linear, quadratic levels and quadratic levels respectively (Table 4). Multiple regression equation generated to predict the melting quality as aff ected by diff erent factors in terms of coded factors is as follows:

Figure 4: Response surface of melting quality as influenced by level of per cent bottle gourd cube and puree.

Melting Quality = + 4.30 - 0.017A- 0.062B- 0.17AB- 0.070A² + 0.045B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.05) negative eff ect on melting quality was observed with rate of puree at linear level (B) and rate of cubes (P<0.05) at quadratic level (A²) respectively. A non-signifi cant negative eff ect on melting quality was observed with cube at linear level (A) while positive eff ect was observed with puree at quadratic level (B²). Th e negative interaction indicates that both the ingredients are negatively correlated and by increasing the level of one ingredient, the level of other ingredient has to be decreased. A signifi cant (P<0.01) negative interactive eff ect on melting quality was observed with rate of cube and puree addition (AB) in ice-cream.

Total score is that property of food which is associated with the acceptance or rejection of product in terms of like. Th e Total score of bottle gourd ice-cream ranged from 84.374 to 91.336 and the bottle gourd ice-cream prepared by using 7% cube and 1.171% puree addition scored lowest total score (Table 3 and Figure 5). Bottle gourd ice-cream prepared by using 7% cube and 4% puree addition was rated best for total score by the panelists. Th e coeffi cient of determination (R²) is the proportion of variability in the data explained or accounted for the model and high value (0.9182) of R2 indicate a better fi t of the model to the data. Also the adequate precision value of 11.056 recommends the use of this response to navigate the design. Th e P value of rate of puree and cube addition showed its signifi cance at linear, quadratic level and at quadratic levels respectively (Table 4). Multiple regression equation generated to predict the total score as aff ected by diff erent factors in terms of coded factors is as follows:

Figure 5: Response surface of flavour as influenced by level of per cent bottle gourd cube and puree.

Total score = + 90.53+ 0.62A + 0.80B-1.56AB-1.31A² - 2.17B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.01) negative eff ect on total score was observed with rate of puree and cube at quadratic level (B² and A²) respectively. A signifi cant (P<0.05) positive eff ect on total score was observed with puree at linear level (B). A non-signifi cant positive eff ect on total score was observed with rate of cube addition in ice-cream at linear level (A). Th e negative interaction indicates that both the ingredients are negatively correlated and by increasing the level of one ingredient, the level of other ingredient has to be decreased. A signifi cant (P<0.01) negative interactive eff ect on total score was observed with rate of cube and puree addition (AB) in ice-cream. No data are reported in the literature on eff ect of addition of processed bottle gourd puree and cubes on acceptability of ice-cream (Table 5).

Table 5: Coeffi cients of Selected Models for Sensory Attributes of Bottle gourd Ice-cream.

Eff ect of diff erent levels of bottle gourd cube and puree addition on physicochemical properties of bottle gourd ice-cream

Physico-chemical attributes play a signifi cant role in deciding the acceptability of the product in new product development process in terms of composition of fi nal product. Th e physicochemical properties chosen for bottle gourd ice-cream were fat, total solids, melting rate, ash and acidity. Th e values of these responses along with their factors are given in Table 6. Th e quadratic model for physico-chemical parameters were obtained through successive regression analysis (Table 7). As seen in Table 8, the calculated F values more than the Table F values at 1% and 5% level of confi dence and it indicated the signifi cance of model terms. Furthermore, the coeffi cient of determination (R²) that refl ects the proportion of variability in data explained or accounted by the model for fat, total solids, melting rate, ash and acidity were 0.9534, 0.9474, 0.8795, 0.9484 and 0.9431 respectively. A larger R² values suggest a better fi t of the quadratic model. Th e adequate precision value of all the parameters studied were greater than 4, hence supporting the suitability of model to navigate the design.

Table 6: Experimental Design Matrix and Scores for Physico-chemical Properties of Bottle gourd Ice-cream

Fat content of ice-cream greatly infl uences the physico-chemical and sensory properties of ice-cream. Fat imparts rich fl avor, soft body and smooth texture and also important in acceptance of ice-cream in terms of consumer’s sensory perception. Th e values of fat in bottle gourd ice-cream ranged from 10.536 to 11.117% of bottle gourd icecream. Bottle gourd ice-cream prepared by using 6% cube and 2% puree addition was having highest% fat (11.12%) content (Table 6 and Figure 6). Th e P values of rate of puree and cube addition showed its signifi cance at linear level (Table 7). Multiple regression equation generated to predict the% fat content in bottle gourd ice-cream aff ected by diff erent factors in terms of coded factors is as follows:

Figure 6: Response surface of fat content as influenced by level of per cent bottle gourd cube and puree.

Table 7: P Values and Partial Coeffi cients of Regression Equation of Suggested Models for Physico-chemical Properties of Bottle gourdIce-cream.

Fat = +10.86 - 0.084A- 0.19B+ 0.028AB- 0.033A² - 4.999E - 003B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.01) negative eff ect on fat was observed with rate of puree (B) and cube (A) at linear while non-signifi cant negative eff ect on fat was observed with rate of puree (B²) and cube (A²) at quadratic level respectively. A non-signifi cant positive interactive eff ect on fat was observed with rate of cube and puree addition (AB) in ice-cream. Th e dilution eff ect of puree in ice-cream was found to exercise signifi cant infl uence (P<0.01) on the fat content. Th ere was a progressive decrease in fat content of experimental samples with increase in level of addition of bottle gourd puree and cubes. However, the average values of fat in all the experimental ice-creams were well above the minimum values specifi ed for ice-cream either legally [34] or as recommended by Bureau of Indian Standards [40] for India.

Th e values of total solid content ranged from 36.601 to 38.202% of bottle gourd ice-cream mix. Bottle gourd ice-cream prepared by using 6% cube and 2% puree addition was having lowest% total solid content (Table 6 and Figure 7). Bottle gourd ice-cream prepared by using 8% cube and 6% puree addition was having highest% total solid content. Th e P values of rate of puree and cube addition showed its signifi cance at linear level (Table 7). Multiple regression equation generated to predict the total solid content in bottle gourd ice-cream aff ected by diff erent factors in terms of coded factors is as follows:

Figure 7: Response surface of total solids content as influenced by level of per cent bottle gourd cube and puree.

Total Solids = + 37.62 + 0.25A + 0.43B- 0.11AB- 0.011A² - 0.063B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.01) positive eff ect on% total solid content was observed with both rate of cube (A) and puree (B) at linear level. Th e negative sign in interaction indicates that both the ingredients are negatively correlated and by increasing the level of one ingredient, the level of other ingredient has to be decreased. A non-signifi cant negative interaction eff ect on% total solid content was observed with rate of cube and puree addition (AB). Th e square of factor indicates that the eff ect of ingredient at highest (quadratic) level used in the product standardization. A nonsignifi cant negative eff ect on% total solid content was observed with both rate of cube (A2) and puree (B2) at quadratic level in bottle gourd ice-cream.

Th e addition of sugar treated processed bottle gourd puree and cubes was found to exercise signifi cant infl uence (P<0.01) on the total solids content of experimental ice-cream. Th ere was a progressive increase in total solids content of experimental ice-cream with increase in level of addition of bottle gourd puree and cubes. Th is eff ect could be attributed to the higher total solid content of bottle gourd puree (57.32%) and bottle gourd cubes (78.86%) compared to total solids content of basic ice-cream mix i.e. 37.37%. Th e average values of total solids content all the experimental samples were well above the minimum values specifi ed for ice-cream either legally [34] or as recommended by Bureau of Indian Standards [40] for India.

A desirable melting quality in a frozen product means the melted product should have the appearance and consistency similar to original mix. Furthermore, meltdown is an important part of any consumer’s perception of product quality. Meltdown behavior provides insights into key ice cream phenomena, such as protein stability, fat agglomeration and air cell size, which infl uence body, texture, and other sensory characteristics. Th e values of melting quality ranged from 53.693 to 57.784% of bottle gourd ice-cream (Figure 8). Th e P values of rate of puree and cube addition showed its signifi cance at linear level (Table 7). Multiple regression equation generated to predict the melting rate in bottle gourd ice-cream aff ected by diff erent factors in terms of coded factors is as follows:

Figure 8: Response surface of melting rate as influenced by level of per cent bottle gourd cube and puree.

Melting Rate = + 57.23- 0.90^*A- 0.73^*B+ 0.61^*AB- 0.98^*A² - 0.94^*B²

Where, A and B refer to% cube and% puree respectively. A signifi cant negative eff ect on melting rate was observed with both rate of cube (A) (P<0.01) and puree (B) (P<0.05) at linear level. Th e positive sign in interaction indicates that both the ingredients are positively correlated and by increasing the level of both ingredients at linear level, it positively infl uence melting rate means melting rate increases. A nonsignifi cant positive interaction eff ect on melting rate was observed with rate of cube and puree addition (AB). Th e square of factor indicates that the eff ect of ingredient at highest (quadratic) level used in the product standardization. A signifi cant (P<0.01) negative eff ect on melting rate was observed with both the rate of cube (A²) and puree (B²) at quadratic level in bottle gourd ice-cream. It means at both linear and quadratic level of both the form of bottle gourd addition decreases melting rate of bottle gourd ice-cream.

Th e values of ash content ranged from 0.857 to 0.897% of bottle gourd ice-cream. Bottle gourd ice-cream prepared by using 6% cube and 2% puree addition was having highest% ash content (Table 6 and Figure 9). Bottle gourd ice-cream prepared by using 8% cube and 6% puree addition was having lowest% ash content. Th e P values of rate of puree and cube addition showed its signifi cance at linear level (Table 7). Multiple regression equation generated to predict the ash content in bottle gourd ice-cream aff ected by diff erent factors in terms of coded factors is as follows:

Figure 9: Response surface of ash content as influenced by level of per cent bottle gourd cube and puree.

Ash = + 0.87 - 6.971E - 003^*A- 0.013^*B+ 9.687E - 004^*AB+1.152E - 003^*A² + 2.017E - 003^*B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.01) negative eff ect on ash was observed with both rate of cube (A) and puree (B) at linear level. Th e positive sign in interaction indicates that both the ingredients are positively correlated and by increasing the level of both ingredients at linear level, it positively infl uences ash content. A non-signifi cant positive interaction eff ect on ash was observed with rate of cube and puree addition (AB). Th e square of factor indicates that the eff ect of ingredient at highest (quadratic) level used in the product standardization. A non-signifi cant positive eff ect on ash was observed with both the rate of cube (A²) and puree (B²) at quadratic level in bottle gourd ice-cream.

Th e values of acidity ranged from 0.1902 to 0.1933% lactic acid of bottle gourd ice-cream. Bottle gourd ice-cream prepared by using 6% cube and 2% puree addition was having highest% acidity content. Bottle gourd ice-cream prepared by using 8% cube and 6% puree addition was having lowest% acidity content (Table 6 and Figure 10). Th e P values of rate of puree and cube addition showed its signifi cance at linear level (Table 7). Th e negative sign indicates that increase in level of cube and puree addition at linear level decreases analyzed% acidity content of icecream. Negative sign in interaction of rate of cube and puree addition in ice-cream decreases% acidity content. Multiple regression equation generated to predict the acidity in bottle gourd ice-cream aff ected by diff erent factors in terms of coded factors is as follows:

Figure 10: Response surface of acidity (% l.a.) as influenced by level of per cent bottle gourd cube and puree.

Acidity = + 0.19 - 5.242E - 004^*A-1.018E - 003^*B- 6.598E - 006^*AB- 2.279E - 004^*A² -1.892E - 004^*B²

Where, A and B refer to% cube and% puree respectively. A signifi cant (P<0.01) negative eff ect on% acidity content was observed with both rate of cube (A) and puree (B) at linear level. Th e negative sign in interaction indicates that both the ingredients are negatively correlated and by increasing the level of one ingredient, the level of other ingredient has to be decreased. A non-signifi cant negative interaction eff ect on% acidity content was observed with rate of cube and puree addition (AB). Th e square of factor indicates that the eff ect of ingredient at highest (quadratic) level used in the product standardization. A nonsignifi cant negative eff ect on% acidity content was observed with both rate of cube (A²) and puree (B²) at quadratic level in bottle gourd icecream (Table 8).

Table 8: Coeffi cients of Selected Models for Physico-chemical Attributes of Bottle gourd Ice-cream.

No data are reported in the literature on eff ect of addition of processed bottle gourd cubes and puree on fat, total solids, melting rate, ash and acidity of ice-cream for comparison

Optimization of Product Formulations

Optimization was done to achieve the best possible combination of% rate of bottle gourd cube and puree addition in ice-cream that would lead to the most acceptable product in terms of sensory scores and physico-chemical attributes. Th e specifi ed selected range was used for obtaining the best optimized combination is illustrated in Tables 9 and 10. Th e data were analyzed in Design Expert Package 9.0.2. Considering the constraints and their limits, the RSM suggested the one most suited solution. Th e fi nal product was made with this formulation and compared with that of predicted values (Table 11). Th e calculated values of ‘t’ for all the parameters being less than the table values, there was no signifi cant diff erence between the predicted and actual values of responses (Table 12). It is, therefore, confi rmed that the selected combination is the best one in terms of the sensory, compositional and physicochemical responses delineated at the beginning of the study.

Table 9: Criteria Chosen For Optimization Process.

Table 10: Suggested Solution from RSM Analysis for Bottle gourd Ice-cream.

Table 11: Predicted Sensory Scores of Optimized Bottle gourd Ice-cream from RSM.

Table 12: Comparison of Predicted V/S Actual Values of Responses.

Physico-chemical, sensory and microbiological analysis of bottle gourd ice-cream manufactured by optimized combination of bottle gourd cubes and puree

Th e approximate composition, physicochemical, rheological, sensory and microbiological quality of the optimized product is presented in Table 13. It was envisaged that the developed bottle gourd ice cream would be high in fi bre content since it has been reported that the fi bre content of bottle gourd (Pusa Naveen) is 0.7% [34]. However, the crude fi bre content of developed bottle gourd ice cream was found to be only 0.05%. Th is could be attributed to use of peeled and de-cored bottle gourd for preparation of bottle gourd puree and bottle gourd cubes. It has been reported that most of fi bre content in bottle gourd is in peel and core portion [34]. Since vegetables are a good source of β-carotene, the β-carotene content was analyzed and was found to be 0.126 mg per 100 g. Th e β-carotene content of control vanilla ice-cream was found to be 0.115 mg per 100 g. Addition of bottle gourd did not result in increasing β-carotene content of bottle gourd ice-cream. Th e Vitamin C content of peeled raw bottle gourd was 12 mg per 100 g. Processed bottle gourd was added in bottle gourd ice-cream at the rate of 10.36%. Th e Vitamin C content was found to be only 0.68 mg per 100 g. Hence, there was about 50% loss of Vitamin C during processing of bottle gourd. Th e calcium, potassium and phosphorus content of raw peeled bottle-gourd were found to be 0.03, 3.35 and 0.5 mg/100 g edible portion respectively. Th e calcium potassium and phosphorus content of bottle gourd ice cream was 0.14, 0.5 and 0.16 g/100 g respectively.

Table 13: Average proximate chemical composition, physicochemical properties and microbiological quality of Bottle gourd Ice-cream manufactured by standardized process.

Selecting compatible fl avor/s in formulation of bottle gourd ice cream and assessing the cost eff ectiveness of the formulated bottle gourd ice cream against vanilla ice cream

To select the most preferred fl avor in bottle gourd ice cream, three fl avors viz. cardamom (fresh powder), vanilla and saff ron were selected. Vanilla and saff ron were added at the rate of 1.5 ml and cardamom powder was added at the rate of 0.15 g per kg of ice-cream mix. Bottle gourd ice cream prepared using above three fl avors was compared with sample of vanilla ice cream. For preparation of bottle gourd ice cream the composition of basic mix was 12.0% fat, 11.0% MSNF, 14.0% sucrose, 0.10% sodium alginate, 0.05% guar gum, 0.02% carrageenan and 0.20% GMS. Th e optimized rate of bottle gourd forms viz. puree and cubes was added at the rate of 4.71 and 6.85% by weight of mix. Th e eff ect of use of diff erent fl avoring ingredients on sensory attributes of bottle gourd ice cream and vanilla ice cream is presented in Table 14.

Table 14: Effect of different fl avouring on sensory attributes of bottle gourd ice-cream and regular vanilla ice-cream.

It can be seen from Table 14 that there was no signifi cant (P>0.05) diff erence in color and appearance and melting quality scores of all the experimental samples. Th e fl avor score of bottle gourd ice cream fl avored with cardamom and vanilla was at par (P>0.05) with vanilla ice cream whereas the fl avor score of saff ron fl avored bottle gourd ice cream was signifi cantly (P<0.05) lower than all the experimental samples. Th e body and texture scored of all the three bottle gourd ice creams were signifi cantly (P< 0.05) lower than sample of vanilla ice cream. Th is could be attributed to lower fat content in bottle gourd ice cream viz. 10.95% versus 12.0% in sample of vanilla ice cream. However, the total score of bottle gourd ice cream fl avored with cardamom was at par (P>0.05) with sample of vanilla ice cream.

Th us, it can be concluded on the basis of this study that among three varieties of bottle gourd available in market Pusa Naveen is most suitable for manufacture of bottle gourd ice cream. A combination of sugar treated bottle gourd puree and bottle gourd cubes added at the rate of 4.71% and 6.85% resulted in most acceptable bottle gourd ice cream. Cardamom was the most preferred fl avor for bottle gourd ice cream. Th e acceptability of developed cardamom fl avored bottle gourd ice cream was at par (P>0.05) with vanilla ice cream.