Chemical characterization of almond meal as a co-product of the mechanical extraction of almond oil

Almonds harvesting, along with precleaning, hulling, shelling, and sorting, generate recoverable co-products such as double, malformed, or damaged almonds. The common way that generates more benefit is the extraction of almond oil for food and cosmetic uses. The aim of this study is to compare composition and nutritional value of almond meals of the main varieties from eastern Morocco, Marcona (M), Fournat de Breznaud (FNB), FerragnesandFerraduel (FF) and Beldi a local ecotype (B), during three consecutive crop years (2016-2018). Significant differences were observed, which mainly concern the residual oil in almond cakes and especially their fibres, sugars, and proteins. Differences depend on the quality of the raw material and the effects of the crop year on the variety. Besides, from a nutritional point of view, analysed almond meals still contain significant amounts of residual oil and the major constituents of almond seeds (sugars, fibres, and proteins). The amino acid profile presents eight essential amino acids (Cysteine, Methionine, Isoleucine, Leucine, Phenylalanine, Tyrosine, and Lysine Valine) and the Protein Digestibility Corrected Amino Acid Score (PDCAAS) ranges between 24.78% for (M) and 62.17% for (FF). We conclude the analysed almond meals present interesting nutritional values as a specific ingredient for gluten-free culinary preparations and light foodstuffs.


Introduction
Almonds are the edible seeds of the Prunus amygdalus Dulcis tree, more commonly known as the almond tree, which is native from its wild ancestors of the central plateau of Asia, currently domesticated, and widely cultivated in Mediterranean climate areas [1]. Although almonds are commonly referred to as nuts, they are actually the seeds found at the centre of the almond fruit. Almond fruit consists of four parts: kernel or Seed or "meat", seed coat (a thin leathery layer known as brown skin of the seed), almond shell, and almond hull which corresponds to the outer green shell cover [2].
The kernel or seed is the edible part of the nut, with a high nutritional value. Almond kernels can be consumed raw or cooked, blanched, ground into flour and even made into non-dairy milk. It can be also incorporated, as an ingredient, into culinary recipes. They are very rich in fat, making them a perfect source of edible oil and oil for cosmetic [3]. In eastern Morocco, thanks to the Moroccan green plan and the PROFAO (Solidarity farming Project of Almond Sector in Eastern Morocco, 2011-2018); 6000 hectares of new almond orchards were created. PROFAO, as an integrated project, allows the organisation of almond's farmer cooperatives and facilitates the creation of economic interest grouping for almond post-harvest facilities, processing, production and marketing of almonds and derivatives [4].
Almond nuts crushing, shelling and sorting operations generate co-products like doubles malformed almonds or mechanically damaged (split and broken kernels) and defects seeds. The best way for exploiting this co-product and deriving more benefits and add value is the mechanical extraction of almond oil for food and cosmetic uses. This process of almond oil extraction also generates another by-product (Almond residue or almond meal) which represents at least 50% by gross mass of the raw material of damaged almonds. Thus, the aim of this study is to assess the nutritional value of these almond residues by analysing the chemical composition of almond meals from fourintroduced varieties, Marcona (M),

Plant material and sampling
In eastern Morocco, producers and processors of almonds in the Sidi Bouhria region are organised into an economic interest grouping (GIE-Sidi Bouhria). In order to derive maximum benefit from their almond production with the perspective of zero waste and using everything the almonds orchard grows. Thus, the almond kernels grading and sorting co-product, corresponding to split & broken malformed and double almonds, is intended for the extraction of almond oil for food and cosmetics. In addition to almond oil, the mechanical process produces a residue "the almond cake /"meal". This study concerns the chemical characterisation of the almond meals of the five most cultivated varieties in the region of eastern Morocco. The almonds come from five-year-old almond trees, installed on a limestone soil located in Sidi Bouhria, in the east of Morocco (GPS coordinates: 34° 44ʹ13.6ʹ'N, 002° 20ʹ15.0ʹ' W). The orchards are rain-fed with a minimum of additional irrigation under semi-arid climatic conditions ( Table 1). The GIE-Sidi-Bouhria provided the almond meals used in this study.

Mechanical extraction of almond oil
The mechanical extraction of almond oil was carried out by a mechanical press system (IBG MonfortsOekotec GmbH & Co.KG, made in Germany). The screw speed was 60 RPM, and the fine particles were removed from the almond oil by filtration followed by centrifugation at 3000 RPM for 15min. The GIE-Sidi Bouhria provided almond meal samples for the Physico-chemical analyses. Samples were crushed and sieved through a 1 mm mesh sieve, then stored at 4°C until use. Analyses were made in triplicate for the almond meal of each variety.

Dry matter
Dry matterwas carried out in accordance with the thermogravimetric method [5]. It consists to dry 5 to 20 g of the homogenized sample at 103° ± 2° C until obtaining a constant weight.

Residual oil
Residual oil in almond meal of deferent varieties was performed by chemical extraction using a Soxhlet apparatus, where 200 ml of hexane was used for a total oil extraction of 5g of powdered almond meal then the solvent was evaporated under vacuum using a rotary evaporator and the oil was collected [6].

Total sugars
Total sugars contentsis determined according to the Bertrand method with little modifications in four main steps:The first step performed by acid hydrolysis and neutralization of the entire hydrolysate to transform all the polysaccharides into simple reducing sugars. The second step is a sample deproteinisation. The third step consistinto precipitation and filtrationof sugars extract. Finally, titration with potassium permanganate was carried out for determination of total sugar contents in analysed samples of almond meal. The result is expressed as a percentage of mass of glucose in 100 g of sample [7].

Dietary fibres
The determination of dietary fibreswas carried out according to the AOAC 991 method [8]. Three successive enzymatic attacks were carried out (α-amylase, protease and αamyloglucosidase) to eliminate starch and proteins. The fibresare precipitated by technical ethanol. The obtained value was corrected by determining the residual ash (Muffle furnace), proteins (Kjeldahl) and by the blank.

Ash
Ash is determined according to the AACC 2001method [9]. The aim is to determine the mineral content of the sample by heating and so destroying the organic material. Thus, 5 g of the already dried sample was introduced in a pre-tared crucible. Then the sample was introduced into a Muffle furnace at 550 °C overnight, then allowed to cool in a desiccator and finally the weight difference was measured.

Proteins
Total nitrogen content was estimated by the Kjeldahl method [10], then protein content was calculated by multiplying the total nitrogen content by a conversion factor of 5.18 [11].

Amino acid profiles
The determination of amino acids by HPLC is carried out according to the method of Spackman et al., [12] by acid hydrolysis of almond meal proteins. This analysis depends on the protein content in sample (500 mg for a sample containing 2% of nitrogen). In our case, 130 mg of each almond meal sample was taken and dissolved in 10 ml of 6 N-HCl containing 0.1% phenol. The samples were mixed and sealed in the presence of nitrogen, and then were heated in oven at 110° C for 24 hours. After cooling in ice, a solution of 30 ml of a citrate buffer at pH 2.2 was added, then a pH adjustment was carried out with 7.5 N and 1 N of NaOH respectively in order to obtain a pH of 2.2. The solutions thus obtained were diluted with citrate buffer (pH = 2.2), after adding 1 ml of Norleucine at 50 μM / ml as internal standard. The solutions were partially filtered by 0.2 µm filter. A 20-µl of aliquot was injected into the HPLC using a sodium oxidase column, cation exchange resin, a column post for the derivation of amino acids into ninhydrin and a UV detector at 570 nm except for Proline at one length 440 nm wave.

Protein quality calculations
The protein quality evaluation was studied by calculation of sum essential amino acid (ΣEAA), sum of non-essential amino acid (ΣNEAA) (indicate the importance of the nutritional value of proteins), Chemical Index (CI) and Proteins Digestibility of Corrected Amino Acid Score (PDCAAS). This later has been adopted as the chosen method for the measurement of the protein value in human nutrition [13]. The Chemical Index (CI) was calculated on the basis of the procedure described previously by Sujak et al (2006),based on following calculation[(% limiting amino acid) sample /(%limiting amino acid) protein of reference )]x100 [14]. In our case, limiting amino acidswere Met + Cys and the protein referenceused of amino acids forhuman food protein composition appropriate for an adult human was FAO/WHO/UNU requirements(2007) ( Table 2) [13].

Statistical analysis
The chemical characterization of almond meal obtained from four varieties and one local ecotype, during three consecutive crop years, was carried out on triplicate for each sample and each parameter. Statistical analyses were conducted using Statistical Package for the Social Sciences (SPSS for Windows, version 21, SPSS Inc., Chicago, IL, USA). The normal distribution was verified according to Shapiro Wilk test. Two-way ANOVA statistical analysis and Tuckey's post-hoc test was used for means comparison, the difference was considered significant (at P<0.05).

Results and discussion
Almond processing coproducts might represent a precious source of raw materials for food, cosmetic, and pharmaceutical industries, however, their valorisation needs a fine physicochemical characterization for the development of novel procedures to be implemented to the sectorial industries. Indeed, providing innovative products from Almond pressed-oil residues (communally named almond oilcake or almond meal) could provide benefit to the overall almond production. Table 3 (table 3) concern three consecutive crop years and six analysed parameters of Almond meals: dry matter, residual oil, sugars, fibres,ash andproteins contents.In general, the chemical composition of almond oil meal change between varieties and crop years mainly for dry matter, sugars and ash.

Dry matter
Dry matter data registered for almond meals (Table 3), range between a minimum of 89.29% for FF-CY1 and a maximum of 94.56 % for M-CY2. Indeed, dry matter values for analysed almond meals of each varieties are slightly different; these differences seem to be mainly related to the harvest year, the variety, and probably to the technique of post-harvest drying. The ANOVA analysis (Table 3) confirms this observation and shows a significant difference at P<0.05. In general, the results are similar to the literature, for instance, Arbouche R (2013) found 94.6% dry matter in the almond meal [17], which is very similar to our findings for M-CY2 (94.56 %). These results are comparable to those of several other types of meals (oil cakes of soybean, groundnut, coconut, cottonseed, sunflower, sesame), whose dry matter contents vary between 85% to 94% [18].

Residual oil
Oil cakes/meals are solid residues, obtained after the extraction of oil from the plant part such as seeds or nuts. In general, residual oil in meals depends on quality of the raw material and on the processing of oil extraction. In spite of the fact that the process used for extraction of almond oils is the same for all varieties, the quantities of residual oil extracted from different almond meals analysed (Table 3) (Table 4) confirms this observation and shows a significant difference at P<0.001. Since the press system is the same, the difference should be related to the raw material, particularly to the physical state of the kernels, moisture and kernels defects but also to the almonds variety [19]. In this regard, Rabadan et al (2017) reported comparable results that range from 15.24% and 25.5 % of residual oil in almond meals [20] and his concluded that residual oil of meals depends on efficiency of mechanical oil extraction. Besides, Savoire et al. (2013) reported that residual oil yield was influenced by many parameters such as kernels variety; press system parameters and pre-treatments preceded this process [21].

Sugars
Whatever the crop year,almond oil cake of the variety Marcona (M) had the lowest sugar content (Table 3)respectively for the three studied years (12.60%; 12.57%; 15.02%), while the variety Fournat(FNB) seems to present the highest sugar content (13.84 /21.65 % / 19.21%). The sugar contents of almond oil meals from the three other varieties are between these limit values. It is certainly here genotype-related differences, which are confirmed by ANOVA analysis ( Table 4) that shows significant differences between genotypes at P<0.05.
The almond oil cake, although considered simply as a by-product of almond oil extraction, is an excellent source of nutrients such as proteins, vitamins, minerals, fibres and bioactive components that may offer a number of health benefits. In this regard recently, in a research collaboration, but also from the literature [22][23][24][25], we show that high phenolic content remains in this coproduct making it usable for improvement of nutritional and functional properties of certain food products. In addition, this coproduct contains significant amounts of other phytochemicals compounds including tocopherols, flavonoids, tannins, and can be used for cosmetics and pharmaceutical purposes.
In the present study, analyses focus on proteins content, amino acids profile ash and dietary fibres.

Dietary fibers and Ash
Dietary fibres can act by changing the nature of the contents of the gastrointestinal tract and by changing how other nutrients and chemicals are absorbed. Fibres and minerals bring nutritional benefits to consumers. The results of the analysed almond cakes for the dietary fibres and ash contents are presented in Table 3. They vary between a minimum of 31.52% /DM registered for Marcona (M-CY2) and a maximum of 42.3%/ DM for Beldi ecotype (B-CY2). The calculated average value of dietary fibres content in analysed almond meals is 35.5%/ DM for the three crop years of this study, this value is in accordance with the literature. Analysis of ANOVA (Table 4) show no significant difference for the factor Year, whereas the factors Variety and Year * Variety interaction present a significant difference between the samples at P<0.05. Finally, the fibres' content of analysed almond meals is clearly higher than those given, by the literature, such as those of sunflower seeds (28.41%) and soybean (17.8%) [19].  Ash contents of the analysed almond meals from the three crop years (Table 3) varied between a minimum of 4.75 % DM for (M-CY3) and 8.15 % DM for (FNB-CY2). Analysis of ANOVA (Table 4) shows that there is a significant difference between the analysed samples at P<0.05.However, the observed differences can not be related to almond genotype or crop year effect. Almond meal is rich in macro elements, namely, potassium, phosphorus and magnesium (data not shown) which make it an important source of these essential elementsRabadan et al. (2017) [20]. In this regard, Karaman et al (2015), recommend, whenever possible, to incorporate a low dose (1 to 2%) of almond flour in food products (for instance baked goods) to improve their nutritional quality and enrich their mineral content [24].

Protein content and amino acid profiles of almond oil cake
The growing interest in almonds and derivatives is due to their nutritious ingredients such as lipids, proteins, carbohydrates, vitamins and minerals, as well as secondary metabolites.
Most studies focus on almond oil, for food uses and especially in cosmetics, that is more profitable, however, few studies concern the protein fraction, which after almond oil extraction constitutes the major component of the almond cake / meal. According to US department of agriculture (USDA) database, the protein content ranges from 16.0 to 22.0% of kernel weight amongst different varieties. Therefore, theoretically after extraction of the oil, which constitutes about 50% of the weight of the kernel, protein contents must double in almond cake / meal. Indeed, protein content of the analysed almond meals from the three cropyears (Table 3) [20] for almond meals of Ferragnes (40.50%) and Ferraduel (44.85%) varieties, but a high protein content (55.93%) for almond meal of Marcona, which is known for its low oil richness. ANOVA test (Table 4) shows a significant difference between the analysed almond meal samples. The significant differences observed seem to be due to both the genetic factor (almond variety) and the environment (crop year) with significant effects respectively at P<0.01 and at P<0.001, whereas the factor "Crop Year * Variety" interaction present a highly significant effect at P<0.001.
Amino acid (AA)profiles ofanalysed almond cake / mealarepresented in Table 5. conclude that the genotype (almond variety) strongly influences the quality (AA profiles) and the protein richness of almond seeds and consequently those of almond's oil cakes. From a nutritional point of view, the analysed almond meal samples show interesting protein values and a medium to pretty good amino acid profile, which is deficient in certain essential amino acids. This finding is in agreement with the results reported by Ahrens et al (2005) [16]. The mean values of chemical indices (CI) and PDCAAS (Proteins Digestibility of Corrected Amino Acid Score) are shown in Table 6. Like all vegetable proteins, which often have one or more limiting amino acids, the analysed samples of almond cakes have relatively low chemical indices (CI),which vary between a minimum of 28.22% for (M-CY2) and a maximum value of 70.81% (FF-CY3). These "CI" values are comparable to those of proteins of vegetable origin and which are generally lower than the CI of proteins of animal origin. The gap between CI values may be due to the genetic and environmental factors, which strongly influence the amino acid profiles in plants [11].All the calculated PDCAAS values for the analysed samples of almond meals are less than 100 but close to reported values for proteins of maize, rice and wheat [21]. "PDCAAS values" for almond meals range between 24.78% for (M-CY2) and 62.17% for (FF-CY3 Finally, we concluded that almond meal coproducts are interesting sources of essential amino acids (EAA, Table 6). They are therefore new good sustainable nutritional alternatives to animal proteins.

Conclusion
This study focused on the characteristics of oil cake as an extraction residue from almond oils and showed, in view of their major constituents (sugars, proteins, fibres, and minerals), that they can be used as an ingredient or food additive, therefore, with an economic advantage for producers, but also from a zero-waste perspective of great environmental interest.
Although this research was more oriented to almonds harvesting and post-harvesting processing stages, the impact of the genotype and the crop year effect, seem very visible, since the quality of the harvest almonds and their coproducts differ from year to year. This depends on the impact of agricultural conditions and interactions between almond varietiesand pedo-climatic factors. Thus, for the studied oilcake samples of different almond varieties and except their fibre contents, significant differences were observed for all the other analysed parameters (Residual oil, Dry matter, Sugars, Proteins, Ash). The chemical composition, of the residues of mechanically pressed almond oil (almond meal), depends mainly on the variety of almond that, in turn, is affected by soil and climatic conditions, which determines the quality of the raw material used for the almond oil extraction.
Moreover, the almond meal coproducts contain beneficial nutrients and bioactive molecules; they constitute an interesting source of protein (EAA), dietary fibres and minerals; they have also the advantage to be a gluten-free product. Therefore, these coproducts of almond oil extraction would be an interesting ingredient for specific gluten-free culinary preparations and in the formulation of light foodstuffs. Indeed, world protein demand is growing strongly and it seems necessary to offer new sustainable and good nutritional alternatives to animal protein. Legume cakes, such as soybean and rapeseed meals, because of their relatively balanced essential amino acid profile (EAA), are widely used in the formulation of animal feed and vegan food. Almond oil cakes, despite their relatively unbalanced essential amino acid (EAA) profile, have the advantage of being E3S Web of Conferences 183, 04004 (2020) I2CNP 2020 https://doi.org/10.1051/e3sconf/202018304004 gluten-free and could be a new ingredient for the formulation of new light foodstuffs and vegan foods.