Application of imaging techniques to evaluate micro-structural and compositional modification during Barley malting

Collins, Carla (2019-12)

Thesis (MScFoodSc)--Stellenbosch University, 2019.

Thesis

ENGLISH ABSTRACT: Over time, barley (Hordeum vulgare) has been domesticated from its wild relative, Hordeum spontaneum to cultivate breeds specialised for malting. The malting industry relies on good quality grain, with good malting characteristics from specialized breeding programmes. These malting characteristics may be expressed in a myriad of ways – most notably in malting barley cultivars, the evidence lies within the starchy endosperm. The endosperm consists of a protein matrix embedded with starch granules. Often, the structure of the endosperm may be described as either mealy or steely based on the arrangement of starch granules and proportions within the matrix. The endosperm of a mealy barley grain has a loosely packed structure with several air spaces between starch granules, while also having a larger portion of A-type starch granules compared to B-type starch granules. An endosperm that is described as steely has a structure that is tightly packed with small, B-type starch granules filling many of the voids to result in a dense endosperm. Due to the arrangement of the endosperm, water uptake into a barley grain may be influenced during the malting process. It has been theorised that mealy barley grain with a less dense endosperm structure allows water uptake to occur with ease, compared to steely barley grain where hydration may be hindered. Obstructed water uptake and uneven distribution of water throughout the barley grain in turn limits the activation of hydrolytic enzymes and subsequent malt modification. With the aid of scanning electron microscopy (SEM), the endosperm structure of barley grain may be closely inspected to evaluate the properties of a good and a poor malting cultivar. Three conventional fixation methods were compared to establish a technique to observe the endosperm structures with minimal artefacts. It was found that plunge freezing and cracking open of grain was an effective method in preserving the structure of the endosperm, compared to chemical fixation that delivered artefacts and cryo-sectioning that caused damage to structures. To observe the endosperm matrix throughout the malting process, plunge freezing was done during successive days of germination and evidence of starch degradation by enzymatic activity was seen in situ. Using the method of plunge freezing, 11 local malting barley cultivars with varying brewing performances were compared to identify characteristics that may predict its brewing performance. Evidence was found that the endosperm of a good converter appears to have a more loosely packed structure with accompanying air spaces, while the endosperm of a poor converter has characteristics of a steely endosperm that slows water uptake. The fact that water has good absorbance properties in the near infrared (NIR) region, lends to the advantage of NIR applications for the study of food products that have a moderate to high water content. Applications of NIR have evolved to incorporate hyperspectral imaging (HSI) that allows specialised applications in the field of cereal grain quality control. As barley grain hydrate during the malting process, modes of water uptake may differ between a good and a poor converter. This theory was investigated using NIR-HSI (in the wavelength range between 850 and 2500 nm) for barley breeders to identify good malting cultivars for the brewing industry. In order to track hydration of barley grain, multivariate image analysis was employed and score plots were recoloured successively in order to track changes that occur over the course of barley grain hydration. Using partial least squares discriminant analysis (PLS-DA), good converting malts were distinguished from poor converting malts at key timepoints along the steeping process. The mean spectra of individual grains were assigned to a most probable class above 0.50 probability. Classification error before steeping was high (27.1%) and gradually lowered to give the most accurate classification model (classification error of 10.4%) at 37 hours of steeping.

AFRIKAANSE OPSOMMING: Gars (Hordeum vulgare) het mettertyd vanuit die verwante wilde Hordeum spontaneum getem geraak en as gespesialiseerde garssoorte vir moutery gekultiveer geraak. Die moutbedryf maak op goeiekwaliteit-graan met goeie mouteienskappe van gespesialiseerde teelprogramme staat. Hierdie mouteienskappe kan op verskeie maniere manifesteer - in moutgarskultivars is die getuienis veral in die styselagtige endosperm te sien. Die endosperm bestaan uit ‘n proteienmatriks en styselkorrels. Die struktuur van die endosperm kan as melerig of staalagtig beskryf word gebaseer op die rangskikking van die styselkorrels en verhoudings binne die matriks. Die endosperm van ‘n melerige garskorrel het ‘n losgepakte struktuur met verskeie lugruimtes tussen die styselkorrels, terwyl dit ook ‘n groter proporsie A-tipe-styselkorrels as B-tipe-styselkorrels het. ‘n Endosperm wat as staalagtig beskryf kan word, het ‘n struktuur van styfgepakte klein B-tipe-styselkorrels wat baie ruimtes vul en ‘n digte endosperm tot gevolg het. Vanwee die plasing van die endosperm kan ‘n garskorrel se wateropname gedurende die moutproses beinvloed word. Daar is geteoretiseer dat melerige garskorrels met ‘n minder digte endospermstruktuur wateropname vergemaklik, in teenstelling met staalagtige garskorrels waar vogopname belemmer word. Belemmerde wateropname en ongelyke verspreiding van water deur die garskorrel lei op sy beurt tot die beperkte aktivering van hidrolitiese ensieme en gevolglike moutmodifikasie. Die endospermstruktuur van garskorrels kan van naderby geinspekteer word met behulp van skanderingselektronmikroskopie (SEM) om die eienskappe van ‘n goeie en ‘n slegte moutkultivar te evalueer. Drie konvensionele fikseringsmetodes is vergelyk om ‘n tegniek te vind om die endospermstrukture met minimale artefakte te beskou. Daar is gevind dat dompelingsbevriesing en oopkraak van graan ‘n effektiewe metode was om die endospermstruktuur te behou, teenoor chemiese fiksering wat artefakte opgelewer het en kriosnitte wat beskadiging van strukture veroorsaak het. Ten einde die endospermmatriks gedurende die moutproses waar te neem is dompelingsbevriesing gedurende opeenvolgende dae van ontkieming gedoen, en aanduidings van styseldegradasie deur ensiematiese aktiwiteit is in situ gesien. Deur die gebruik van dompelingsbevriesing is 11 plaaslike moutgarskultivars met wisselende broudoeltreffendheid vergelyk om eienskappe te identifiseer wat hul broudoeltreffendheid vooraf kan bepaal. Getuienis is gevind dat die endosperm van ‘n goeie omskakelaar ‘n meer losgepakte struktuur met gepaard gaande lugruimtes het, teenoor die endosperm van ‘n swak omskakelaar wat eienskappe van ‘n staalagtige endosperm het wat wateropname vertraag. Die feit dat water goeie absorbansie-eienskappe in die naby-infrarooi (NIR) omgewing het, verleen die voordeel van naby-infrarooi (NIR) toepassings vir die studie van voedselprodukte wat ‘n gemiddelde tot hoe waterinhoud het. Naby-infrarooi (NIR) toepassings het ontwikkel tot die inkorporering van hiperspektrale beeldanalise (HSI) wat gespesialiseerde toepassings in graansoort-kwaliteitsbeheer teweegbring. Wanneer garskorrels vog tydens die moutproses opneem, kan die wyse van wateropname tussen ‘n goeie en ‘n swak omskakelaar verskil. Hierdie teorie is ondersoek deur die gebruik van naby-infrarooi hiperspektrale beeldanalise (NIR-HIS) (in die golflengteveld tussen 850 en 2500 nm) sodat garstelers goeie moutkultivars vir die broubedryf kan identifiseer. Ten einde vogopname van garskorrels te kan volg is meerveranderlike beeldanalise gebruik en hoofkomponent-tellingplotte is agtereenvolgend oorgekleur om die veranderings te kan waarneem. Parsiele kleinstekwadrate-diskriminant-analise (PLS-DA) is gebruik om goeie omskakelingsmoute van swak omskakelingsmoute tydens sleuteltydstippe van die weekproses te kan onderskei. Die gemiddelde spektra van individuele graankorrels is ‘n mees waarskynlike klas van bo 0.50 waarskynlikheid toegeken. Klassifikasiefoute voor weking was hoog (27.1%) en is geleidelik verlaag om die akkuraatste klassifikasiemodel (klassifikasiefout van 10.4%) by 37 uur weking te gee.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/107256
This item appears in the following collections: