Biophysical Control of Coffee Quality: the Case of Southwestern Ethiopia

Biophysical control of coffee quality: the case of southwestern Ethiopia

Kassaye Tolessa

Dutch translation of the title:

Biofysische controle van de koffiekwaliteit: het geval van Zuid-West Ethiopië.

Suggested way of citation:

Kassaye, T. 2017. Biophysical control of coffee quality: the case of southwestern Ethiopia. PhD thesis, Ghent University, Gent, Belgium,

ISBN:978-90-5989-982-7

The author and the promoter give the authorization to consult and to copy parts of this work for personal use only. Every other use is subject to the copyright laws. Permission to reproduce any material contained in this dissertation should be obtained from the author. Biophysical control of coffee quality: the case of southwestern Ethiopia

Kassaye Tolessa

Thesis submitted in fulfilment of the requirement of the degree of Doctor (Ph.D) in Applied Biological Sciences

Department of Applied Analytical and Physical Chemistry Isotope Bioscience Laboratory - ISOFYS Faculty of Bioscience Engineering Ghent University, Gent, Belgium Academic year 2016 – 2017 Promoter

Prof. Dr. ir. Pascal Boeckx Department of Applied Analytical and Physical Chemistry, ISOFYS, Ghent University, Gent, Belgium

Chair of the Examination Committee

Prof. Dr. ir. Patrick Van Damme Department of Plant production, Ghent University, Gent, Belgium

Board of Examiners

Prof. Dr. ir. Geert Haesaert (Secretary) Department of Applied Biosciences, Ghent University, Gent, Belgium

Prof. Olivier Honnay Department of Biology, Ecology, Evolution and Biodiversity Conservation, KU, Leuven

Prof. Dr. ir. Koen Dewettinck Food Safety and Food Quality, Ghent University, Gent, Belgium

Dr. Laurence Jassogne International Institute of Tropical Agriculture, Systems Agronomist, Kampala, Uganda

Dean of the of Faculty of Bioscience Engineering Prof. Dr. ir. Marc Van Meirvenne

Rector of Ghent University Prof. Dr. Anne De Paepe

Acknowledgements

Doing PhD is both painful and enjoyable experience. It is just climbing a high peak mountain step by step; going together with so many people’s kind help. So completion of this thesis would not have been possible without the support and guidance of different people. I would, therefore, like to offer my sincere gratitude to all of them. Firstly, I would like to express my special appreciation and thanks to my esteemed promoter, Prof. Pascal Boeckx for his meticulous guidance, keen interest, encouragement and constructive criticism that have made this PhD study a reality. I greatly appreciate the initiation, encouragement and negotiation he made for NIR machine procurement that I used for my PhD research and it is a great contribution for future coffee quality research in Jimma University and in the country in general. I also want to gratefully thank Prof. Bernard De Baets and Dr. Michael Rademaker for their support and critical guidance on NIR machine and software. My special thanks are also due to Rita and Karoline for hands on training on NIR machine. I also thank Prof. Luc Duchateau for his kind support for bringing coffee samples from Ethiopia to Belgium, data analysis and facilitating the shipment of our NIR machine from Belgium to Jimma, Ethiopia.

My special thanks also go to the entire ISOFYS staff for the nice atmosphere they have and for exemplary experience I got from their day to day activities. I especially would like to thank Katja Van Nieuland for her kind help for NIR procurement, calibration, solving some technical problems and her unforgettable support in coffee grinding. My special gratitude also goes to Saskia Van der Looven for her wonderful room arrangement for my stay in Gent and for all other administrative issues she managed for me. I also thank Dr. Samuel Bodé for his assistance in solving problems I encountered in biochemical composition analysis and other technical issue. It is also a great pleasure for me to thank Prof. Oswald Van Cleemput for his critical comments and suggestions that improved my papers a lot. I would like also to thank Eric Gillis for translating the thesis summery to Dutch. My thanks also go to Prof. Bruno De Meulenaer for allowing me to use the coffee grinding machine, GC and HPLC equipment in his laboratory. I greatly thank Sofie and Nathalie, who were always assisting me in reading GC and HPLC measurements.

I greatly appreciate the support received from my MSc students, Jolien D’heer and Desalegn Alemayehu. I want also to acknowledge expertise in Goma I and Limu kossa coffee plantations who helped me so much during coffee processing and drying. Berhanu

Acknowledgements

Tsegaye and Amsalu Abera receive special thanks for their technical support and relevant information in every up and down. I greatly appreciate the support I got from Adamu Tadesse, Bahilu Ijo, Zakir A/Mecha, Aliye Hussein and Wubit Jorga. Many thanks go to Gerba Daba, Diriba Bane and Talila Geremew for their unreserved help during data collection. I would like to convey my heartfelt thanks to Jimma University, College of Agriculture and Veterinary Medicine (JUCVAM) for offering me ideal environment to concentrate on my own research and PhD study. I give my sincere thanks to all people in JUCAVM. My deep appreciation also goes to sisters and brothers from Evangelical Christian Fellowship, Gent, Belgium and JUCVM.

My very sincere thanks to Dr. Ghislain de Groote and Dr. Bonamie Anne-Marie for their parent-like support, generous care and the home feeling while I am way from home. I will miss you both very much.

I owe a lot to my Mum, who encouraged and helped me at every stage of my personal and academic life and longed to see this achievement come true. I deeply miss my father, my father-in-law and mother-in-law, who are not with me to share this joy. I am very much indebted to my beloved husband, Dr. Adugna Debela (Kennaa tiyyaa) and my lovely and adorable daughters, Sena and Lensa. Being both a father and a mother while I was away was not an easy task for my husband. He took every responsibility to take care of my kids and family. Kennaa tiyyaa, you are a true, great supporter and an exemplary for everyone. Without your kind support and encouragement, motivation and prayers, I would not have finished this work. There are no words to convey how much I love you and thank you so much for everything you did. I want to express my deepest appreciation to Sena and Lensa for their great patience, understandings and for being outstanding students. Senu, I will never forget the letter you wrote me. It has a lot of blessings and encouragement. Those inspirational words gave me strength and motivated me to work harder. Lensu, you were always hoping to come to Belgium via Skype. That is amazing and unforgettable. I am also very grateful to my sisters-in-law, my brothers-in-law, my brothers, my sisters, my uncles and aunts for all of the sacrifices that they have made on my behalf. Very special thanks to Amalework, Ajaíbdu, Derartu and Lomitu for their unreserved help in taking care of my kids and managing everything at home in my absence.

ii Acknowledgements

I am also grateful for the support of my prayer groups: Gash Geremew, Mesfin Getachew, Yonas Simion, Bisrat Getachew, Kabe Gerba, Shimelis Kassaye, Mesert, Alemayehu Getachew, Mitike, Tewdrose and Sara. Your continuous prayer for me was what sustained me thus far. I gratefully acknowledge NUFFIC, Netherlands organization for International Cooperation in Higher Education (NICHE) for financial support. I would like also to express my sincere appreciation to NICHE project coordinators and supporters: Carl Jaspers, Prof. Derbew Belew, Umi Abdulkadir, Getachew Nigussie and Selamawit Teshome. I am also very grateful to Ethiopia Commodity Exchange (ECX) for their support in evaluating coffee cup quality as well as the famers from Mana, Goma and Limu district for allowing us to work on their coffee farms.

Above all, I thank my God, my good Father, for letting me through all the difficulties. I have experienced your guidance day to day. You are the one who let me finish my degree. I will keep on trusting you for my future. Thank so much you, Lord.

iii

Table of contents

Acknowledgements ...... i Table of contents ...... v Summary ...... ix Samenvatting ...... xiii

List of abbreviations ...... xvii

Chapter 1- General introduction ...... 1

1.1 Agriculture and economic development ...... 2

1.2 Coffee tree and its importance ...... 3

1.3 Coffee production systems in Ethiopia ...... 6

1.4 Coffee cherry processing ...... 8

1.5 Coffee actors along the value chain ...... 9

1.6 Coffee bean quality ...... 13

1.7 Need for fast and reliable coffee quality assessment method ...... 17

1.8 Biochemical composition ...... 18

1.9 Factors influence coffee quality and biochemical composition ...... 24

1.10 Objectives and outline of the PhD thesis ...... 32

Chapter 2- Influence of growing altitude, shade and harvest period on quality and biochemical composition of Ethiopian specialty coffee ...... 35

2.1 Introduction ...... 37

2.2 Materials and methods ...... 39

2.2.1 Study site ...... 39

2.2.2 Treatments and experimental design ...... 39

2.2.3 Coffee quality analysis ...... 40

2.2.4 Caffeine and chlorogenic acid extraction ...... 42

2.2.4.1 Liquid chromatography analyses ...... 42

2.2.5 Moisture content determination ...... 43

2.2.6 Data analysis ...... 43

2.3 Results ...... 43

2.3.1 Coffee quality analysis ...... 43 v

Contents

2.3.2 100-bean weight, caffeine and total chlorogenic acid content ...... 47

2.4 Discussion ...... 49

2.5 Conclusions ...... 55

Chapter 3 - Ethiopian specialty coffee quality and biochemical composition in relation to growing altitude and postharvest processing ...... 57

3.1 Introduction ...... 59

3.2 Materials and methods ...... 62

3.2.1 Study site ...... 62

3.2.2 Treatments and experimental design ...... 62

3.2.3 Processing methods ...... 63

3.2.4 Coffee quality analysis ...... 64

3.2.5 Biochemical compostion analysis ...... 64

3.2.6 Moisture content determination ...... 65

3.2.7 Data analysis ...... 65

3.3 Results ...... 65

3.3.1 Coffee quality analysis ...... 65

3.3.2 Caffeine, total chlorogenic acid and sucrose content ...... 68

3.4 Discussion ...... 69

3.5 Conclusions ...... 73

Chapter 4- Analysis of coffee quality along the coffee value chain in Jimma zone, Ethiopia

...... 75

4.1 Introduction ...... 77

4.2 Materials and methods ...... 80

4.2.1 Study site ...... 80

4.2.2 Treatments and experimental design ...... 80

4.2.3 Coffee quality analysis ...... 82

4.2.4 Data analysis ...... 83

4.3 Results ...... 83

4.3.1 Coffee quality ...... 83

Contents

4.4 Discussion ...... 89

4.5 Conclusions ...... 94 Chapter 5- Prediction of specialty coffee cup quality based on near infrared spectra of green coffee beans ...... 97

5.1 Introduction ...... 99

5.2 Materials and methods ...... 101

5.2.1 Description of the study site ...... 101

5.2.2 Coffee samples ...... 101

5.2.3 Coffee quality analysis ...... 102

5.2.4 NIR spectra measurement and analysis ...... 103

5.3 Results ...... 104

5.4 Discussion ...... 109

5.5 Conclusions ...... 110

5.6 Supplementary informations ...... 112

Chapter 6 - General discussion, conclusions and future perspectives ...... 117

6 General discussion ...... 118

6.1 Optimization of biophysical factors to improve coffee bean quality ...... 119

6.2 Coffee quality along the value chain: evidence from farmers, cooperatives and private

traders ...... 121

6.3 Need for improved coffee quality assessment method ...... 123

6.4 Final conclusions ...... 124

6.5 Practcal implications ...... 126 6.6 Perspectives for future research ...... 127

References ...... 129

Appendix ...... 149

Curriculum vitae ...... 153

vii

Summary

Coffee is the world's most popular beverage and the second largest traded commodity after oil. It is appreciated mainly for its characteristic taste and aroma. The crop has been known and remained Ethiopia’s most important source of foreign currency, accounting 25-30 % of GDP. The country, on the other hand, is recognized as the endemic region of origin and wide diversity of Arabica coffee and has enormous, unexplored potential to produce top specialty coffees. The increase in demand for high quality, specialty coffee with specific characteristics triggered coffee producing countries to produce specialty coffees for these high-end markets. For instance, in Ethiopia, using total specialty cup scores, coffees are grouped into two top specialty classes: specialty 1 (Q1, scores ≥85) and specialty 2 (Q2, scores 80 - < 85). This classification resulted in market segmentation and higher prices at international markets. Thus, improving coffee quality is a key opportunity for increasing coffee exports and price for coffee producing countries like Ethiopia. The factors that determine quality along the value chain are numerous. Genetic traits, growing environment, management practice (e.g. harvest period and postharvest processing methods) and actors within the value chain are thought to affect coffee quality. In this PhD thesis, influences of altitude and shade combined with harvest period and postharvest processing methods on quality and biochemical composition of coffee beans in Mana district, Jimma zone, Ethiopia has been investigated. Variability of coffee qualities among different coffee actors along the value chain combined with sorting and postharvest processing methods in Mana, Goma, and Limu districts were also studied. Furthermore, this PhD study developed a near infrared (NIR) based model to predict coffee cup quality from green coffee bean NIR spectra.

Coffee quality increased with increasing altitude and shading. Coffee beans grown at high altitude (1950 – 2100 m asl.) gave superior coffee quality and about 87% of the coffee beans qualified for Q1 coffees. At mid (1600 – 1650 m asl) and low (1450 - 1550 m asl) altitude, however, almost all coffee beans classified as Q2. At high altitude, dense shade (65-85%) decreased the percentage of Q1 coffee, while the largest shares of Q1 coffees (ca. 67%) were obtained from medium shading and open sun (ca. 53% Q1). At mid altitude, on the other hand, medium shading (40-55%) allowed to enhance the share of Q1 slightly from 7 to 13% compared to the open sun and dense shading. The study also showed that harvesting period and coffee processing method affect coffee quality.

ix Summary

Coffee beans harvested at early to mid season period improved coffee bean quality. Compared to late harvesting, early harvesting increased the percentage of Q1 coffee from 27 to 73%, at high, and from 7 to 20% at mid altitude. Moreover, dry processing positively influenced coffee quality and increased the percentage of Q1 coffee beans by more than 50% compared to washed coffee. However, to improve the quality of dry processed coffee beans, proper coffee cherry e.g. ripe red coffee cherries need to be delivered. Sorting coffee cherries before dry processing further enhanced the share of Q1 coffee beans. In addition, because of easy of management, low production cost and eco-friendly nature, dry processing should be recommended over washing methods.

Both altitude and shading also affect biochemical composition of green coffee beans. Coffee beans grown at mid altitude and open sun accumulated higher content of total chlorogenic acid (46.5 g kg-1). At high altitude and open sun, on the other hand, the content decreased to 40.5 g kg-1. The highest caffeine content (17.9 g kg-1) was obtained from mid altitude, dense shade and early harvested coffee beans, whereas the lowest level (14.5 g kg-1) was found at high altitude, medium shade and mid season harvested coffee beans. In addition caffeine and sucrose were significantly affected by interactions between altitude and processing methods. High altitude grown coffee beans processed via dry, semi-washed and washed processing methods gave higher caffeine content than low altitude grown coffee beans. However, washed processed coffee beans gave lower sucrose content to dry and semi-washed beans at high altitude. At mid and low altitude, on the other hand, processing methods did not show significant differences on any biochemical compounds. Moreover, regardless of processing methods, the content of caffeine, total chlorogenic acid and sucrose increased with altitude by 11, 9, and 10%, respectively. In general the study indicates that altitude is the main driving factor affecting the compositions of green coffee beans.

Furthermore, coffee beans managed by cooperatives gave better quality than coffee beans managed by private traders. About 33 and 67% of cooperative coffee beans fell in Q1 and Q2, respectively. The majority of the coffee beans collected from private traders classified mainly into commercial grade 3 (ca. 78%), only 22% of the samples were qualified as Q2 and none as Q1. Proper harvesting and sorting of coffee cherries would help cooperatives and private traders to further enhance the share of speciality coffee from their production.

x Summary

Besides coffee quality improvement, nowadays, the need for a fast and reliable coffee quality assessment method is increasing. Coffee cup quality assessment can be assessed organoleptically by professional coffee tasters, based on established terminologies for cup quality analysis (e.g. flavor, acidity and body). However, these evaluation methods are subjective, costly, time consuming, require sample preparation and may end up in biased grading systems. NIR is an increasingly growing technique for its fast, simple, cheap, reliable and its ability to measure multiple qualities attributes at a time. In this study, a NIR-based model predicting coffee cup quality from green bean NIR spectra has been also successfully developed. For this, partial least square (PLS) regression method was used to develop a model correlating spectral data to cupping score data (i.e. cup quality). The results obtained suggest that NIR spectra of green coffee beans are a promising tool for fast and accurate prediction (i.e. Q1 vs. Q2) of specialty coffee quality. Thus, the method indeed simplifies coffee grading systems and reduces the time needed to grade beans by the conventional method (i.e cup tasting). To achieve this, however, institutional (e.g. ECX) decision to integrate this new method into the current cup tasting system is required. In addition, the model needs to be further tested using a large number of data sets and coffee samples collected from different coffee growing regions in Ethiopia.

All together this PhD identified important biophysical factors that control coffee bean quality along the coffee value chain, opened concepts and methodologies for enhancing coffee bean quality and developed a promising tool for fast and accurate assessment of specialty coffee cup quality.

Samenvatting

Koffie is 's werelds meest populaire drank en de tweede grootste verhandelde grondstof na olie. Het wordt voornamelijk gewaardeerd voor zijn karakteristieke smaak en aroma. Het gewas is en bleef bekend als Ethiopië ’s belangrijkste bron van buitenlandse valuta, goed voor 25-30% van het BBP. Het land, aan de andere kant, wordt erkend als de endemische regio van herkomst en de grote diversiteit aan Arabica koffie en heeft een enorm, onontgonnen potentieel om top koffiespecialiteiten te produceren. De toename van de vraag naar kwalitatief hoogwaardige, speciale koffie met specifieke eigenschappen prikkelde koffieproducerende landen om speciale koffies te produceren voor deze high-end markten. Bijvoorbeeld, in Ethiopië, met behulp van de totale specialiteit cup scores, worden koffies gegroepeerd in twee specialiteit topklassen: specialiteit 1 (Q1, scores ≥85) en specialiteit 2 (Q2, scores 80 - <85). Deze classificatie resulteerde in marktsegmentatie en hogere prijzen op de internationale markten. Zodoende, is het verbeteren van de kwaliteit van de koffie een belangrijke opportuniteit voor het verhogen van de koffie export en de prijs voor koffieproducerende landen zoals Ethiopië.

De factoren die de kwaliteit in de waardeketen bepalen zijn talrijk. Genetische eigenschappen, groeiomgeving, beheerpraktijk (bijv oogstperiode en de naoogst verwerkingsmethoden) en actoren binnen de waardeketen worden verondersteld de kwaliteit van de koffie te beïnvloeden. In dit proefschrift werden invloeden van hoogte en schaduw in combinatie met oogstperiode en de naoogst verwerkingsmethoden op kwaliteit en biochemische samenstelling van koffiebonen in het Mana district, Jimma zone, Ethiopië onderzocht. Variabiliteit van koffie kwaliteiten tussen verschillende koffie actoren in de waardeketen, in combinatie met het sorteren en naoogst verwerkingsmethoden in de districten Mana, Goma en Limu, werden ook bestudeerd. Bovendien ontwikkelde dit doctoraatsonderzoek een nabij-infrarood (NIR) gebaseerd model waarmee koffiekop kwaliteit van groene koffieboon NIR spectra voorspeld kan worden.

De koffiekwaliteit nam toe met toenemende hoogte en schaduw. Koffiebonen, geteeld op grote hoogte (1950 - 2100 m boven zeeniveau) gaven een superieure koffiekwaliteit en ongeveer 87% van de koffiebonen gekwalificeerd voor Q1 koffies. Medio (1600 - 1650 m boven zeeniveau) en lage (1450 - 1550 m boven zeeniveau) hoogte, echter, werden

xiii Samenvatting

bijna alle koffiebonen geclassificeerd als Q2. Op grote hoogte, verlaagde dichte schaduw (65-85%) het percentage Q1 koffie, terwijl het grootste aandeel van Q1 koffie (ca. 67%) werd verkregen door medium beschaduwing en open zon (ca. 53% Q1). Op medio hoogte, anderzijds, liet medium beschaduwing (40-55%) toe het aandeel Q1 enigszins te verbeteren van 7 tot 13%, in vergelijking met de open zon en dichte beschaduwing. De studie toonde ook aan dat de oogstperiode en koffie verwerkingsmethode invloed hebben op de koffiekwaliteit. Koffiebonen geoogst bij vroeg tot midden seizoen periode verbeterde de koffieboonkwaliteit. Vergeleken bij late oogst, verhoogde het vroege oogsten het percentage van Q1 koffie van 27-73%, bij hoge en van 7-20% bij medio hoogte. Bovendien beïnvloede droogbewerking de koffiekwaliteit positief en verhoogde het percentage Q1 koffiebonen met meer dan 50% ten opzichte van gewassen koffie. Echter, om de kwaliteit van droog bewerkte koffiebonen te verbeteren, moeten goede koffiebessen,b.v. rijpe rode koffiebessen, worden geleverd. Het sorteren van koffiebessen, voorafgegaan aan droge verwerking, versterkt verder het aandeel van Q1 koffiebonen. Bovendien, als gevolg van eenvoudig beheer, lage productiekosten en milieuvriendelijk karakter, valt droge verwerking aan te bevelen boven wasmethoden.

Zowel hoogte als beschaduwing beïnvloeden ook de biochemische samenstelling van groene koffiebonen. Koffiebonen geteeld op medio hoogte en in open zon accumuleerden een hoger gehalte aan totale chlorogeenzuur (46,5 g kg-1). Op grote hoogte en in open zon, anderzijds, nam de inhoud af tot 40,5 g kg -1. Het hoogste cafeïnegehalte (17,9 g kg-1) werd verkregen vanaf medio hoogte, dichte schaduw en vroeg geoogste koffiebonen, terwijl het laagste niveau (14,5 g kg-1) werd gevonden op grote hoogte, halfschaduw en in het midden seizoen geoogste koffiebonen. Bovendien werden cafeïne en sucrose significant beïnvloed door interacties tussen hoogte en verwerkingsmethoden. Koffiebonen, geteeld op grote hoogte, verwerkt via droge, semi- gewassen en gewassen verwerkingsmethoden gaven een hoger cafeïnegehalte dan koffiebonen geteeld op lage hoogte. Echter, gewassen verwerkte koffiebonen gaven een lagere sucrosegehalte dan droog en semi-gewassen bonen op grote hoogte. Op medio en lage hoogte, anderzijds, vertoonden de verwerkingsmethoden geen significante verschillen op biochemische verbindingen.

Bovendien, ongeacht de verwerkingsmethoden, verhoogde het gehalte aan cafeïne, het totale chlorogeenzuur en sucrose met de hoogte met 11, 9, en 10%, respectievelijk. In het algemeen geeft de studie aan dat de hoogte de belangrijkste drijvende factor is die xiv Samenvatting

de samenstellingen van groene koffiebonen beïnvloedt. Bovendien gaven koffiebonen, beheerd door coöperaties een betere kwaliteit dan koffiebonen beheerd door particuliere handelaren. Ongeveer 33 en 67% van de coöperatieve koffiebonen vielen in Q1 en Q2, respectievelijk. De meerderheid van de koffiebonen van particuliere handelaren werd voornamelijk ingedeeld in de handelskwaliteit 3 (ca. 78%), slechts 22% van de monsters werden gekwalificeerd als Q2 en geen enkel als Q1. Gepast oogsten en sorteren van koffiebessen zou coöperaties en particuliere handelaren helpen om het aandeel van specialiteit koffie uit hun productie te verbeteren.

List of abbreviations

AOAC Association of Official Analytical Chemists ARDO Agriculture and Rural Development Office AT Aftertaste Ca. Calculated CGA Chlorogenic Acid CFAQ Caffeoylferloyl-qunic acid CQA Caffeoylqunic acid CQIAC Coffee Quality Inspection and Assurance Center diCQA Dicaffeoyquinica acid ECX Ethiopia Commodity Exchange FAO Food and Agricultural Organization FLO Fair-trade Labelling Organization FQA Feruloylquinic acid GC Gas chromatography GDP Gross Domestic Product GI Geographic Indications HPLC High Performance Liquid Chromatography ISO International Coffee Organization JARC Jimma Agricultural Research Center JUCVM Jimma University College of Agriculture and Veterinary Medicine LC-TOF-MS Liquid chromatography-time of flight-mass spectrometry m a.s.l Meter above sea level MFA Ministry of Foreign Affairs NIRS Near-infrared spectroscopy NICHE Netherlands Organization for International Cooperation in Higher Education OCP Overall cup preference PLS Partial Least Square PQ Physical quality PTR Proton Transfer Reaction PTQ Preliminary total quality

xvii List of abbreviations

p-CoQA p-coumaroylquinic acid RPD Residual prediction deviation RMSECV Root mean square error of cross validation RMSEP Root mean square error of prediction SPME Solid-phase microextraction TSCQ Total specialty cup quality UK United Kingdom

xviii Chapter 1

General Introduction

Chapter 1

1. Introduction

1.1 Agriculture and economic development

Progress in fighting against poverty continued worldwide but still about 795 million people are estimated to have been undernourished in periods of 2014 – 2016 (FAO, 2015). The vast majority of these people (ca. 98%) are from developing countries. The progress towards improving food security varies across countries. Countries, for instance, in central Asia, east Asia, Latin America and north Africa showed remarkably rapid progress in reducing hunger while countries in South Asia and Sub-Saharan Africa progress at slow pace (FAO, 2012). Countries in East Africa, in particular, remains with the biggest hunger problem and an estimated 124 million people are undernourished. In most of the developing countries, agriculture is the largest sector creating employment opportunities and playing a major role in supplying food and export earnings for people in the world’s rural areas (Dercon & Gollin, 2014). About 75% of these people are mainly dependent on agriculture. Thus, the major opportunity for poverty reduction and job creation through a “smart” and coordinated approach can be found in the agricultural sector. Hence, to alleviate poverty and support food security in Africa, smallholder agriculture should be the main target (Garvelink et al., 2012).

Ethiopia, with extremely high population growth, is the second largest country in Africa. About 85% of the population are living in the rural areas and depend on agriculture for their livelihood. Agriculture is the main driver of the country's economy and accounts for more than 40% of national gross domestic product (GDP), 90% of exports, and provides basic needs and incomes to more than 90% of the poor people (Diao et al., 2010). Tafesse (2005) also reported that a 1% increase in Ethiopia’s agricultural per capita value, declines the poverty level of rural households by 1%. Ethiopia has been well known for its agricultural development. However, the rapid growth in population decreased land resources and expanded cultivation to forest areas and steep slopes (Diao et al., 2010). These resulted in extreme land shortages especially in the highland part of the country where most population are living and most agricultural products are grown. In recent years, policy reforms and agricultural investments have provided a boost to agricultural production, primarily cereals and other export commodities (World

General introduction

Bank, 2006). The performances of agricultural sector over the last five years e.g. production of market driven products, crop diversifications and increased export commodities, are an indication of the new direction of the country's development strategy. Among the export commodities, coffee is the one with due attention and is the crop earning largest foreign exchange for the country.

1.2 Coffee tree and its economic importance

Coffee belongs to the genus Coffea L. and family Rubiaceae (Berthaud & Charrier, 1988; Cannell, 1985; Coste et al., 1992). The family Rubiaceae has over 6000 species and 500 genera (Davis & Rakotonasolo, 2001). The crop is woody perennial evergreen dicotyledonous shrub that grows to 3 -12 m if not prune. The coffee tree has two distinct types of branches: vertical (orthotropic) and lateral (plagiotropic). As the vertical grows, lateral branches grow from buds produced at leaf axils. The first levels of plagiotropic branches, inserted on an orthotropic parent are called primaries. Primary branches generate from the “head of series” buds (buds located in the leaf axils and give rise to secondary branches, which, in turn, give rise to tertiary and quaternary ramifications (Wintgens, 2012). Flowering occurs at some distance from the apex. On a branch, the zone with flowers and fruits moves in apical direction; at tree level, flowering and fruiting move from lower to higher branches. Flowering is triggered by rain or temperature drops. In a zone with buds ‘ready to flower’ fractions of flower buds may open sequentially upon several inductive events (DaMatta et al. 2007). It requires 3 to 4 years, starting from time of seed germination to flower and fruit bearing. It also requires 7- 8 and 11-12 months for arabica and robusta, respectively from flowering to fruit maturity (Wintgens, 2012).

Coffea is the most economically important genus in the family Rubiaceae (Pearl et al., 2004; Wellman, 1961) and over 100 species have been described within Coffea L. Despite this diversity, only two are grown commercially on a large scale: Coffea arabica L. (Arabica coffee) and Coffea canephora Pierre ex Froehner (Canephora or Robusta coffee), both originate from Africa (Clifford & Wilson, 1985; Lashermes et al., 1999; Raina et al., 1998; Wrigley, 1988). Coffea arabica is the world's most important agricultural commodity, accounting ca. 60%for the world market while Robusta coffee contributes the remaining ca. 40% (ICO, 2016). Coffee Arabica also produces superior quality compared to Robusta coffee, but more susceptible to diseases and pests

Chapter 1

(Agwanda et al., 1997). Coffee Arabica grows in altitudes ranging from 1300 – 2800 meter above sea level with an optimum mean daily temperature of 18 to 21°C (Jha et al., 2014). It also grows under-shade or without shade, from sandy to humic soils and from moist to semi-arid zones (Bayetta, 2001). However, Robusta coffee is grown mainly in tropical lowlands, below 1000 m with mean daily temperature ranging from 24 to 26°C (Ky et al., 2001). In East Africa, Robust coffee is also grown above 1000 meter above sea level. Nowadays coffee trees are grown in tropical and subtropical regions of Central and South America, Africa, and South East Asia, mainly in regions with temperate and humid climates (Schenker, 2000) (Fig.1.1).

Coffee is one of the most popular beverages in the world. It is appreciated largely for its characteristic taste and aroma and more recently, for its beneficial effects on human health (Perrone et al., 2008). Coffee as a beverage also has a strong historical, cultural, social and economic importance. Coffee is grown in more than 80 countries in tropical and subtropical regions of the world and exported in various forms e.g. green or roasted beans to more than 165 countries (Abreu et al., 2012; Prakash et al., 2002). It accounts for 75% of export revenue and provides livelihoods for ca.25 million smallholder coffee producers around the world. According to ICO (2016) statistical report, about 8.9 million ton of green coffee beans were produced in 2015/2016. The share of top ten coffee producing and exporting countries to the world market is given in Figs. 1.2 and 1.3.

Fig.1.1: Map of coffee production and consumption in the world

General introduction

Guatemala Mexico Uganda (2.3%) (1.9 %) (2.7%) Hunduras (3.6 %)

India (3.7 %) Bra zil Ethiopia (32.7%) (4.5 %) Indonesia (7.8 %)

Colombia (9.5 %) Vietnam (18.6%)

Fig. 1.2: Share of top ten coffee producing countries in the world market (ICO, 2016).

Brazil and Vietnam are the leading coffee producers and exporters countries and contributes about 51% to the world coffee market. Ethiopia is the largest coffee producer in Africa with half of its production directed to export and 10th exporters in the world (ICO, 2016). This shows that the country will get more benefit by increasing coffee production volume and improving its quality

Peru Gatemala (2.9%) (2.7%) Ethiopia (2.7%) Uganda (3.0 %)

Hunduras (4.8%) Bra zil (30.3 %) India (5.3%) Indonesia Colombia (5.3%) (11.0%) Vietnam (20.4%)

Fig. 1.3: Share of top ten coffee exporting countries in the world market (ICO, 2016).

Chapter 1

In addition, the country has huge potential as it is endowed with diverse agro ecologies and wide range of coffee varieties, which have diverse quality profiles. Moreover, more than 95 per cent of coffees produced in the country are by default an organic coffee i.e. no artificial fertilizers and pesticides are used during the production. The production system gave Ethiopia a comparative advantage in the international specialty coffee market (Bart et al., 2014). The crop is known as the backbone of the Ethiopian economy, contributing 25-30 per cent of total export earnings and plays a major role in sustaining the livelihoods of more than 15 million coffee growing households in the country (Davis et al., 2012; Tefera & Tefera, 2014).

1.3 Coffee production systems in Ethiopia

Coffee is mainly grown in the western, southwestern, southern and eastern part of Ethiopia (Fig.1.4). Four major coffee production systems (Fig.1.5) are commonly identified based on the differences in intensity of management: forest coffee, semi-forest coffee, garden coffee and plantation coffee (Hylander & Nemomissa, 2008; Labouisse et al., 2008; Mekuria et al., 2004).

Study region

Fig.1.4: Coffee producing regions of Ethiopia. Names of the regions are given in their trade names. The arrow indicate the study region, Jimma ( i.e Djimmah).

General introduction

B B

Fig.1.5: Coffee production systems in Ethiopia: forest (A), semi-forest (B), garden (C) and plantation coffee (D).

In forest coffee, plants are self-sown and grown under natural forest with no or little management (Hylander & Nemomissa, 2008; Labouisse et al., 2008). Coffee beans from these ‘wild coffee plant’ are collected by locals, but have little direct commercial value. However, they represent an important reservoir of genetic diversity of arabica coffee, (Anthony et al., 2002). The wild populations of arabica coffee in the forests are genetically diverse and have large number of endemic species (Gil et al., 2004; Labouisse et al., 2008). Therefore, the forest coffee has an immense importance in conserving coffee gene pool which can lead to improvement of desirable traits in cultivated coffee e.g. higher quality specialty grade (Berecha et al., 2014), low caffeine content (Silvarolla et al., 2004) and resistance to disease and insect pest (Boisseau et al., 2009). Forest coffees are also needed to develop coffee that adapt to changing climate and market demands (Aerts et al., 2017).

Chapter 1

Semi-forest coffee systems are the advanced form of forest coffee. Farmers transplant semi-wild coffee plants that regenerate spontaneously inside the forest to fill open spaces (Hundera et al., 2013; Labouisse et al., 2008). Through selective thinning of developing tree species and saplings, canopy of semi-forest coffee system is more open than forest coffee system (Aerts et al., 2011). The conversion of forest coffee to semi- forest coffee system negatively affects the diversity of tree species. Furthermore, further intensification leads to the degradation of this semi-forest to the so called semi-plantation coffee system, causing a significant tree species diversity losses. Semi-plantation coffee system involves modification of the forest similar to the semi- forest coffee system, but more intensively, including the systematic planting of coffee seedlings (Hundera et al., 2013). De Beenhouwer et al. (2016) also reported the presence of high anthropogenic disturbance in semi-plantation that cause a relatively species poor canopy consisting of tree species compared to semi-forest coffee system. In semi-plantation coffee system mulching is common practice and organic fertilizers are used locally (De Beenhouwer et al., 2016) but artificial fertilizers are also used in the system (De Beenhouwer et al., 2015b). This shows that the existence of wide variation in semi-forest coffee system is based on the degree of management applied to the system.

In garden coffee systems, coffee tree is planted by farmers and grown in homesteads i.e. close to the house under a few shade trees usually combined with other crops and fruits. Mulching and organic fertilizer are applied to improve soil fertility status. In plantation coffee system, coffee is grown on large commercial farms, state owned farms or private company. This system also represents shade plantations with only few, large canopy trees. Shade trees are mainly indigenous e.g. Acacia spp. and C. macrostachyus but there is also a recently, fast-growing exotic species have been introduced e.g. Grevillea robusta. Modern technologies such as pruning, stumping, mulching and use of chemical fertilizers are adopted in order to increase the coffee yield (De Beenhouwer et al., 2016; Labouisse et al., 2008).

1.4 Coffee cherry processing

The process of bringing the harvested coffee fruits to consumers as a beverage involves a series of steps. Greater control of each step of this process improves the ability to produce a good quality coffee. After harvesting, fruits undergo primary processing to

General introduction generate green coffee beans. To produce green coffee beans, there are two commonly applied methods of coffee cherry processing: dry and washed processing methods. Both methods remove the outer layers of coffee cherries (pericarp) and reduce the moisture content in the coffee beans to about 10–12% (Knopp et al., 2006). During dry processing (unwashed or sun-dried coffee), coffee cherries are directly sundried on mats, concrete, or cement floors in layers of approximately 10 cm (Evangelista et al., 2014). The dried coffee cherries are then de-husked using a coffee hulling machine and green beans are ready for marketing. For coffee beans to dry, this process can take several days to a few weeks, depending on the climatological conditions.

In the washed method (or wet coffee), the handpicked red ripe coffee cherries are mechanically de-pulped and a fermentation processes is carried out for approximately 24 – 72 h. (i.e. variable based on the climatic conditions) to remove the mucilage covering coffee beans (Silva et al., 2005). After washing, the coffee beans are sundried on raised beds. In this process, coffee drying occurs quickly because there is no skin or mucilage present. The dried beans with parchment are then de-husked using a coffee hulling machine to obtain green coffee beans. This process uses about 20 litres of water per kilogram of coffee cherries (Mburu et al., 1994; von Enden et al., 2002).

In addition to aforementioned coffee processing methods, semi-washed coffee processing method is another method recently introduced to coffee producing countries. In this method, the fermentation step is omitted and the mucilage is removed by a de- mucilage machine. The method is known for its simplicity and improved water use efficiency compared to fully washed coffee processing method. It uses about 4 litres of water per kilogram of coffee cherries and leads to more than 80% water reduction compared to washed processing method (Bello-Mendoza & Castillo-Rivera, 1998; Deepa et al., 2002; Matos & Monaco, 2001).

1.5 Coffee actors along the value chain

Value chain represent a full range of activities, which are required to bring a product or service passing through different actors to reach the consumers (Kaplinsky & Morris, 2000). Ethiopian coffee value chain involves mainly the following actors: input providers (seeds, fertilizers), farmers, cooperatives, private traders, private company, coffee

Chapter 1 plantation or state owned farms, union, Ethiopia Commodity Exchange (ECX), exporter, Coffee Quality Inspection and Assurance Center (CQIAC) and consumers (Fig.1.6). Majority of coffee produced in the country (95%) is grown by smallholder farmers with an average productivity of 0.7 ton per hectare (Tefera, 2015). The remaining part is supplied via private company and coffee plantations. They produce and process coffee by themselves. Recently, almost all coffee plantations are owned by private investors. Coffees produced by smallholder farmers, however, reach the final consumers by passing through cooperatives or private traders. A cooperative is a group of smallholder coffee farmers, who buy coffee cherries from their respective member of farmers and process the coffee cherries themselves. Cooperatives are responsible for organizing farmers and play an active role in supporting their members by supplying price information, training on coffee production, initial capital and transportation as well as facilities for coffee processing. Generally, the cooperatives are able to overcome any barriers more easily than individual farmers and enable farmers to improve coffee quality and ultimately their income.

A union is an association of cooperative societies and help to increase the bargaining power of cooperatives and respective farmers. Union buy coffee from their respective cooperatives, processes (e.g. removing parchment or sorting) and sell coffee beans on the international market (Kodama, 2007; Petit, 2007). Cooperative unions have two possible channels to sell their coffee beans. As it is shown in figure 4, cooperatives unions can sell their coffee beans directly (C1-channel) or via ECX (C2-channel) to the international buyers (Tefera, 2016). The former improves farmers’ income and also provides the benefit of assuring traceability, which is increasingly demanded in international markets. In general, cooperative offers economic, social and educational benefits for its members and the related societies. This is because the main target of cooperation is improving the members’ living. However, there are still some challenges (e.g. lack of funds, limited management capacity) that limit the capacity of the cooperatives to fully function to the best of their potentials. These need institutional intervention to enhance their capacities and develop clear governmental policies and supportive legal frameworks that the cooperatives follow and administer in their members (Dejen & and Matthews, 2016; Kodama, 2007).

General introduction

The other coffee market chain is from farmers to private traders. Majority of the farmers, non-cooperatives members, sell their coffee to private traders. Private traders are those traders that have licences to buy coffee cherries from farmers and process, de-hull and export the coffee beans via exporters. Private company or growers, on the other hand, can produce, processes, de-hull and export their own coffee. Both private traders and private company can sell their coffee to exporters only through ECX. Similar to coffee cooperatives, coffee beans produced in coffee plantations or state owned farms can be sold directly to international buyer (P1- channel) or to exporter through ECX auction market (P2 - channel) (Figure 4). In general coffee beans that meet export standard can be exported while under grade coffee beans are used for domestic consumption.

Ethiopia Commodity Exchange (ECX) is a recent initiative for Ethiopia and creates a new marketplace for all market actors in 2008. Subsequently, a law was passed that fully prohibited coffee trading outside the ECX. It aims to eliminate the huge number of middlemen involved in coffee supply chain to reduce transaction cost. This enables coffee farmers to benefit from prevailing market prices. As the result farmers are selling their coffee cherries directly to cooperatives and private traders.

Coffee quality test and grading are also conducted at ECX laboratories. Currently, there are nine ECX coffee cupping laboratory/centres located in different region: In southwest (Jimma, Bedelle and Bonga), south (Awassa, Dilla and Sodo), west (Gimbi), east (Dire Dawa) and one in Addis Ababa. There are three to four Q-certified cuppers for quality testing and grading of coffee in each ECX.

All coffee actors’ i.e. private traders, cooperatives, private company and coffee plantation can deliver their coffee to ECX for quality analysis and grading. The detail coffee quality analysis and grading system is presented below in section 1.6. Once coffee is grade and traded at ECX, exporters pick up their coffee from regional ECX warehouses and do further processing e.g. sorting to improve the quality. Then the quality of the coffee beans will be analyzed and graded again to ensure as it meets export standards by CQIAC in Addis Ababa. For coffee that meet export standard a quality certificate is issued by the CQIA and attached to the coffee beans to be exported. This export standard mainly refers to the quality requirement for export which set by ECX (John, 2011).

Chapter 1

Farmers Private company Coffee plantations

Red Red/dry Red/dry Red/dry cherry cherry cherry cherry

Cooperatives Private traders

Processing: Processing: Processing: Processing: x Pulping x Pulping x Pulping x Pulping /washing /washing /washing /hulling /washing /hulling

C2 ECX warehouse: x Quality analysis C1 x Grading Under grade P x Trading 2

Union Exporter

Processing: P1 Hulling /sorting Processing: Hulling /sorting

Coffee Quality Inspection and Assurance Center: quality analysis, grading and certification Domestic

consumption Export

Fig.1.6: Overview of Ethiopian coffee value chain 12

1.6 Coffee bean quality

Awareness of quality, nutrition and health aspects of coffee among consumers is increasing demand for high quality, speciality coffees (Barbosa et al., 2014; Cheng et al., 2016; Upadhyay & Mohan Rao, 2013). This growing in demand and consumption of high quality coffee with specific characteristics results in market segmentation (e.g. specialty coffee market) and creates a strong potential and new opportunities for coffee-producing countries (Läderach et al., 2011; Oberthür et al., 2012; Wood, 2015). Market segmentation for specialty coffee attracts new customers and results in higher coffee prices (Labouisse et al., 2008). For instance, specialty coffee beans receive a premium price of ca. 20-50% compared to regular coffee beans. Certification also raises coffee prices (e.g. organic and fair trade coffee generate premium of about ca. 9%). In general the share of speciality and certified coffee are estimated to 20 and 12 %, respectively, while the remaining 68 % is sold, mainly locally, as bulk coffee (Bart et al., 2014).

In order to meet consumers’ needs national (e.g. trade mark) and international policies (e.g. food safety, traceability, certification) laying down strict quality requirements. This helps to guarantee the production of higher quality products e.g. origin-linked quality and geographical indications (GI) to promote, protect and value quality of coffee. This also gives consumers further information on what they buy (Curzi & Pacca, 2015). This increase in attention paid to quality and the growing set of regulations in developed countries has also increased the pressure on producers from developing countries to produce coffee beans with specific origin and specific taste (Jouanjean, 2012). Thus, improving the quality of exported coffee represent a necessary pre-condition for economic growth and development of countries like Ethiopia (Curzi et al., 2014).

Ethiopia is naturally gifted with a suitable climate and has the potential to produce single origin specialty Arabica coffee beans with a wide range of flavours (Coste et al., 1992; Labouisse et al., 2008). Premium price (ca. $7- 8 per kilogram of coffee beans) has been paid for specialty coffee of some origin such as Harar, Yirgacheffe and Sidama. Thus, an increased specialty coffee market is thought to be beneficial for Ethiopia to remain competitive in the international market provided the country produces high quality coffee and supply remains stable. However, majority of coffee quality in the country does not meet the international market requirement. For example, the coffees produced in Jimma

Chapter 1 zone, southwestern Ethiopia general fell in lower grade and receive lower prices compared to other regions (Tsegaye et al., 2014). For the last six years, the percentage of specialty coffee 1(Q1) and specialty 2(Q2) supplied by producers from the zone were ca.5% and 12%, respectively. However, this area is known as centre of coffee diversity and high production potential contributing 27% of the country’s coffee export (Dessie, 2008). Several research studies have been done to enhance coffee productivity via improved variety and agronomic practice (Bayetta, 2001; Kufa & Burkhardt, 2011; Kufa et al., 2007). However, quality focused studies were not well articulated and documented. These show that much effort is needed to improve Ethiopian coffee quality so that the country remains competitive in the international market.

Coffee bean quality is mainly described by physical quality, cup quality (sensory quality attributes) and chemical bean constituents (Cheng et al., 2016; Fridell, 2014; Santos et al., 2012).

Physical quality analysis

Physical quality refers to colour, size and shape of beans and presence of defects found in certain coffee batches. The defects include sour beans, black beans, immature or floaters, foreign materials (e.g. stones), broken or nipped, mouldy and insect damage (John, 2011; LMC, 2006). These defects can be associated with problems that occur during growing environment, harvesting, processing, drying and storages of coffee beans resulting in cup quality deterioration (Toledo et al., 2016). For instance, sour beans can be due to lack of water during fruit development and to inappropriate fermentation of beans (Farah, 2012). Sour beans may also result in the formation of black beans (Mazzafera, 1999; Pimenta, 2003; Toledo et al., 2016). Black beans are also the result of carbohydrate deficiency (Illy & Viani, 2005) or microbial fermentation of beans while still on the tree or during postharvest processing (Farah, 2012). In Ethiopia, this coffee bean defect is mainly due to a biotic stress, inappropriate harvesting (immature and overripe) and processing (under or over fermentation). The physical quality assessment is mainly done via visual inspection and gives scores between 0 and 40.

General introduction

Sensory quality analysis

Sensory analysis is a multidisciplinary area comprising measurement, interpretation and understanding of human responses to product quality as perceived by the senses such smell and taste (Singham et al., 2015). The sensory analysis of a product's quality is of great importance when it comes to the decision of whether a consumer will buy it again, or if a different product will be chosen (Van Oirschot & Tomlins, 2002). Sensory analysis has a capacity to predict the response of human to product ingredient, processing and shelf life, to determine whether a new product is worth launching onto the market, for flavour profiling to identify the sensory attributes driving consumer preferences. Thus, sensory analysis is the only method that provides integrated, direct measurements of perceived intensities of sensory quality attributes (Bleibaum et al., 2002).

Some attempts are being made to mimic human sensory systems using instrumental methods such as electronic devices (e-nose and e-tongues). E-nose and electronic tongue (e-tongue) are sensor technology that attempts to mimic the function of human senses. These devices composed of a chemical sensor arrays, coupled with an appropriate pattern recognition systems able to interpret complex signals from such sensors and produce a fingerprint of the product (Ferreira et al., 2007; Wilson & Baietto, 2009). E-nose has been used to analyze the aromatic compounds of several food and beverages. For instance, for evaluation of coffee cup quality and the detection of defects of the beans (Michishita et al., 2010; Rodríguez et al., 2009).

E-tongue is a device that able to recognize tastes or to give information about groups of substances present in complex system (Kobayashi et al., 2010). It uses a range of sensors that respond to salts, acids, sugars and bitter compounds (Baldwin et al., 2011). This approach has been already used for analyzing coffee taste profiles (Fukunaga et al., 1996; Vlasov & Legin, 1998) and recognized as a promising chemical sensors for coffee quality analysis. But some studies revealed that there is still lack of correlation between instruments data with cup quality data since these devices are sensitive to temperature (Baldwin et al., 2011). It has been also shown that the combination of these electronic devices and Gas chromatography–mass spectrometry (GC-MS) instruments are more suitable to analyze specific aromatic compounds (Ongo et al., 2012).

Chapter 1

Moreover, proton transfer reaction (PTR)-MS technique is a rapidly growing analytical technique for detecting and identifying very small quantities of chemical compounds in headspace (Ellis & Mayhew, 2013). For instance, a robust and reproducible model was developed to predict the cup quality attributes of espresso coffee using PTR-MS. In other studies, solid-phase microextraction (SPME)-GC-MS has been proposed to predict coffee cup quality attributes e.g. acidity and flavour based on the chromatographic profiles of volatile compounds of roasted coffee beans (Bressanello et al., 2017; Ribeiro et al., 2012; Ribeiro et al., 2009).

The sensory quality of coffee is the brewed quality or cup quality of the coffee which determine the desirability of coffees for consumption and acts as a yardstick for price setting (Roche, 1995; Walyaro, 1983). Cup quality characteristics are attributes of coffee that can be distinguished by senses and assessed organoleptically by professional coffee tasters, trained cup tasters based on established terminologies for cup quality analysis e.g. acidity, body, flavour and a cup cleanness attributes. The sum of these four cup quality attributes gives preliminary cup quality with a score between 0 and 60. Hence, the sum of both physical and preliminary cup quality gives preliminary total quality with scores ranges between 0 and 100 and is used to classify the coffee beans into different grades (grade 1 = 91 - 100, grade 2 = 81 - 90, grade 3 = 71 - 80, grade 4 = 63 - 70, grade 5 = 58 - 62, grade 6 = 50 - 57, grade 7= 40 - 49, grade 8 = 31 - 39, grade 9 = 20 - 30 and under grade (grade 10) = 15-19 (Table 1.1). Coffee grades ranging from 1 to 3 qualifies for specialty coffee and further assessed based on overall cup preference, acidity, body, aroma, flavour, aftertaste, uniformity, cup cleanness, sweetness and balance each on a scale ranging from 0 to 10. The sum of these cup quality attributes gives total specialty cup quality. Based on the total specialty coffee scores, samples will be further grouped into specialty 1 (Q1) ≥ 85, specialty 2 (Q2) (80 - < 85). The preliminary total quality assessment scores for grade 2 should be higher than 80 to be qualified as specialty 1 Q1coffee (Lemma, 2015).

In Ethiopia, a panel of three cup tasters is commonly used to evaluate the coffee cup quality at ECX cupping centers. Five cups of brewed coffee are considered to evaluate one coffee sample and each cupper taste and give score for each of the five cups. The cupping score variability between cuppers is lower than 1 point on a 0 – 100 scale (Worku et al., 2016).

General introduction

Table 1.1: Overview of Ethiopian Commodity Exchange (ECX) preliminary quality parameters and contribution to a total score of 100 points Preliminary quality analysis Grade categories and total points Grade Total points Physical quality (40) Primary defect (10) Grade 1 91-100 Secondary defects (10) Grade 2 81-90 Shape and make (10) Grade 3 71-80 Odour (5) Grade 4 63-70 Colour (5) Grade 5 58-62 Grade 6 50-57 Grade 7 40-49 Grade 8 31-39 Cup quality (60) Cup cleanness (15) Grade 9 20-30 Acidity (15) Under grade 15-19 Body (15) Flavour (15) Preliminary total (100) Grade 1-3 qualifies for specialty coffee quality Source: John (2011)

Table 1.2: Overview of ECX specialty cup quality parameters, and contribution to a total score of 100 points and specialty coffee categories Specialty cup quality attributes Specialty coffee categories and points Grade Total points Cup cleanness (10) Specialty 1 (Q1) 85-100 Acidity (10) Body (10) Specialty 2 (Q2) 80 - < 85 Flavour (10) Aroma (10) Commercial (Grade 3) 71- < 80 Aftertaste (10) Uniformity (10) Sweetness (10) Balance (10) Overall cup preference (10)

Total specialty cup quality (100) Source: John (2011).

1.7 Need for fast and reliable coffee quality assessment method

As explained above in section 1.6 coffee cup quality analysis in Ethiopia and elsewhere is done via cup tasting which are however, more or less subjective, requires detailed sample preparation e.g. roasting, grinding, brewing, time consuming and expensive. It is also difficult to implement when many samples need to be analyzed during peak coffee

Chapter 1 production. Although the analytical methods e.g. GC-MS and PTR-MS provide reliable and accurate quality profile they are expensive and time consuming. This may limit their application for routine analysis in quality control. Therefore, investigating fast and robust analytical alternatives tools are of great interest.

Nowadays, the coffee industry has a growing need for a rapid, low-cost, reliable and more reproducible analytical method as an alternative tool to evaluate coffee quality attributes and, eventually, other chemical compounds presents in coffee beans. Recently near-infra spectroscopy (NIRS) has emerged as a reliable and promising tool for objective assessment of coffee quality attributes and chemical composition e.g. caffeine, chlorogenic acid content. It also has a potential to measure multiple chemical constituents simultaneously avoiding extensive sample preparation (Barbin et al., 2014; Krähmer et al., 2015). This could bring economical benefits to the coffee sectors and increase consumer confidence in the quality of products. It has been used for the discrimination of Arabica coffee from Robusta (Buratti et al., 2015; Esteban-Díez et al., 2007), defective from non-defective coffee beans (Craig et al., 2012; Craig et al., 2015) and quantification of coffee bean chemical compounds (Huck et al., 2005; Zhang et al., 2013). In other studies e.g. Ribeiro et al. (2011), reported the potential of NIRS for predicting coffees cup quality attributes from roasted coffee beans. However, the possibility of NIRS based coffee cup quality prediction from green coffee beans is not yet investigated. This requires due attention as predicting coffee cup quality from green bean is simpler and cheaper than roasted coffee beans. Hence, investigating the possibility of NIRS approach coffee quality evaluation from spectra of green beans seems to be a key research topic.

1.8 Biochemical composition

The chemistry of coffee quality is highly complex with a wide range of compounds that found in green coffee beans. The key compounds are caffeine (0.9 to 3%), chlorogenic acids (CGA, 5 to 8%) and sucrose (7 to 11%) (De Castro & Marraccini, 2006; Ky et al., 2001). These chemical constitutes are regarded as precursors for coffee cup quality attributes e.g. aroma and flavour (Cheng et al., 2016; Joët et al., 2010). Roasting is the most important factor in development of complex flavour and aroma that make coffee enjoyable (Toledo et al., 2016; Wei et al., 2013). In roasted coffee beans, about 900

General introduction volatile compounds were identified e.g. furans, pyrazines, ketones, pyrroles and alcohols. But only about 5% of these compounds are actually involved in aroma and flavor formation (Flament, 2002; Yeretzian et al., 2003). In this PhD research we focused on sucrose, caffeine and chlorogenic acid as main chemical composition because these attributes are the dominant precursors for coffee aroma and flavour development. Recently, caffeine and chlorogenic acid have also been the subject of several investigations in view of their positive effects in humans health (Duarte et al., 2010).

Caffeine

Caffeine ( 1,3,7-trimethyl-xanthine) is a purine alkaloid and secondary metabolite of the coffee plant where it acts as a natural pesticide (Ashihara & Crozier, 2001). Theobromine and xanthine are key intermediates in caffeine biosynthesis (Ashihara et al., 2008). The major route to caffeine is xanthosine → 7-methylxanthosine → 7-methylxanthine → theobromine → caffeine. Caffeine is formed in leaves, immature coffee fruits and gradually accumulates during fruit maturation. The transcription and enzyme activity in caffeine biosynthesis is high in immature fruits and decreases drastically in the last stages of bean development (Cheng et al., 2016; Perrois et al., 2015). This suggesting that the transcriptional control of caffeine-related genes have considerable effects on caffeine accumulation and are reduced by 50% at the end of bean maturation (Perrois et al., 2015).

Caffeine is not highly transported within the plant. The largest percentage of caffeine remains in the tissues where it was synthesized (Mazzafera & Gonçalves, 1999). In coffee beans, caffeine content during maturation, nearing 1% on a dry matter basis at the time of harvest. The major increase in caffeine content in seeds occurs during the time of rapid endosperm expansion (De Castro & Marraccini, 2006). Apart from these synthesis aspects of caffeine in developing fruits, the catabolism of caffeine influence the caffeine content of beans (Sridevi & Parvatam, 2014). Vitoria and Mazzafera (1997) also reported that at fruit ripening stage, ratio between biosynthesis and biodegradation controls the difference in caffeine production.

Caffeine is a non volatile compound and as the result the content is not affected by roasting (Oestreich-Janzen, 2010). Caffeine is also responsible for the bitter

Chapter 1 characteristic of coffee beans contributing to coffee quality. Recently, caffeine is known as one of the most known behaviorally active drug in the world and mainly consumed from coffee (Davis et al., 2006; Oestreich-Janzen, 2010). Positive health effect of moderate caffeine consumption are also reported in some studies e.g. against cardiovascular and Alzheimer’s disease (Nehlig, 2012; O'Keefe et al., 2013; Tunson, 2014). It has been also reported that caffeine increases energy availability, alertness and decreases fatigue and boosts physical performance (Pesta et al., 2013). However, too much caffeine may result in undesired effects such as cardiovascular disease, depression, and even addiction(Jiang et al., 2014).

Chlorogenic acid

Chlorogenic acids (CGA) are a family of polyphenolic compounds, which are esters between trans-cinnamic acid (such as caffeic, ferulic and coumaric acid) and quinic acid (Stalmach et al., 2010; Trugo & Macrae, 1984).There are three major subgroups of CGA found in coffee: Caffeoylquinic acids (CQA) with 3 isomers (3-, 4- and 5-CQA), dicaffeoylquinic acids (diCQA) (3,4-diCQA; 3,5-diCQA; 4,5-diCQA) and feruloylquinic acids (FQA) (3-, 4- and 5- FQA), p-coumaroylquinic acids (pCoQA), with 3 isomers, (3-, 4- and 5- pCoQA), and six mixed diesters of caffeoylferuloyl-quinic acids (CFAQ) (Clifford, 2000; Clifford et al., 2003).

Similar to caffeine biosynthesis, the major chlorogenic acid (CGA) class, 5-CQA, di3, 5CQA accumulate primarily in immature coffee beans and decrease dramatically during maturation. For instance, di3, 5-CQA decreased from 8.4% in green coffee beans to 2.3% in red mature coffee beans. This is because of some enzymes which are activated at the early bean development at 90-120 days after flowering to produce the major components and this accumulation reached a plateau at 120 to 150 days after flowering. At later stage, from 150-210 days after flowering, the remobilization of chlorogenic acid occur for lignin biosynthesis, responsible for cell wall hardening (Joët et al., 2010).

Chlorogenic acid is one of the chemical constitute that determines coffee flavor and acidity (Trugo & Macrae, 1984). It is also known for its dietary antioxidant effects (Clifford, 1999; Olthof et al., 2001; Stalmach et al., 2010; Zhao et al., 2012). From epidemiology investigations, a health benefit of chlorogenic acid has been demonstrated by several

General introduction studies. For instance, chlorogenic acids reduce risks of type 2 diabetes (Salazar- Martinez et al., 2004; Van Dam & Hu, 2005), obesity (Thom, 2007), high blood pressure and stroke (Larsson et al., 2011; Zhao et al., 2012), anti-hypertensive (Kozuma et al., 2005; Ochiai et al., 2004) and act as plant protection against pathogen infection (Niggeweg et al., 2004; Peterson et al., 2005). Other studies also indicated that consumption of chlorogenic acids contributes to insulin action to enhance glucose tolerance and lipid metabolism (Cho et al., 2010; de Sotillo et al., 2006).

Sucrose

Sucrose is one of the most abundant simple carbohydrates present in green coffee beans. It contributes to more than 90% of the total low molecular carbohydrates in green coffee beans (Knopp et al., 2006). In early developmental stages of the perisperm to endosperm transition, sucrose biosynthesis remains constant. Genes regulating sucrose metabolism are expressed differently during bean development. Metabolism increases during endosperm development stage and dramatically increases during bean color change (from green to yellow) but later it slows down until maturity (Joët et al., 2009). Higher sucrose synthase and acid invertase activities are found to be expressed in early development stages than at the end of bean development stage (Privat et al., 2008). Sucrose acts as aroma precursors, originating several substances (e.g. furans, aldehydes and carboxylic acids) that will affect both flavour and aroma of the beverage (Farah et al., 2006). Sucrose also reacts with amino acids to produce pyrazines and carbonyl compounds, which are involved in flavour formation (Feldman et al., 1969). The ratio of sucrose to amino acid in green beans determines the profile of volatile compounds that might explain coffee quality attributes (Tressl et al., 1982). The highest sucrose content in green beans could be related to good cup quality coffee (Ky et al., 2001; Toci & Farah, 2010).

Coffee bean roasting

Coffee flavor and aroma are largely related to the volatiles compounds produced during roasting. Roasting is the important step in the production of coffee in which coffee beans undergo a series of reactions leading to several changes in chemical composition (Franca et al., 2009; Ribeiro et al., 2011). The major biochemical compounds,

Chapter 1 responsible for the volatiles in roasted coffee are trigonelline, chlorogenic acids, sucrose, lipids and proteins (Ribeiro et al., 2009). Temperature and time of roasting will significantly influence the development of flavour and aroma compounds (Madihah et al., 2013).

The roasting process can be divided into three phase: i) an initial drying phase, evidently endothermic, during which moisture is eliminated. The smell of the beans changes from green to bread-like, and the colour turns to yellowish, ii) the actual roasting phase, during which a number of complex pyrolytic reactions take place. The chemical composition of the beans is drastically modified, with release of large amounts of carbon dioxide and the formation of the many hundreds of substances associated with coffee aroma and taste. The beans change colour from red to a dark brown. Pyrolytic reactions reach a maximum at 190–210°C, when the process becomes endothermic with the release of volatile compounds. and iii) a final rapid cooling phase to stop the final exothermic part of the roasting operation, using cooling agent (e.g. air) (Buffo & Cardelli -Freire, 2004).

The principal reactions occur during roasting process are Maillard, Strecker reactions and other thermal reactions. As a result of these chemical reactions, aromatics, acids and other flavour components are created, balanced or altered in a way to build the perfect flavour, acidity and aroma of the coffee beverage. The Maillard reaction (non- enzymatic browning) occurs between a reducing sugar and an amino acid or a protein, which is responsible for flavour and colour development. Hundreds of coffee flavour compounds are formed from Maillard chemistry, including the potent coffee aroma flavor compound, 2-furfurylthiol. Strecker degradation is a chemical reaction between an amino acid and α-dicarbonyl with the formation of an aminoketone and an aldehyde. A schematic representation of the reactions occurring during the roasting process is given in Fig.1.7.

So far, more than 800 different compounds from a wide range of chemical classes have been identified in roasted coffee but only about 25-35 are considered as key odorants, responsible for the coffee flavour (Flamen, 1989; Sberveglieri et al., 2012). Some of these volatile compounds are furans and pyrans, pyrazines, pyrroles, ketones and phenols, hydrocarbons, alcohols, aldehydes, acids and anhydrides, esters, lactones, thiophenes, oxazoles, thiazoles, pyridines, amines, and various sulfur and nitrogen

General introduction compounds (Toci & Farah, 2008). Pyrazines and aldehydes are among the two most important categories formed during roasting that contribute to a desirable coffee flavour (Madihah et al., 2013; Makri et al., 2011; Toci & Farah, 2008). Optimization of the roasting conditions (e.g. temperature and time) is very crucial in the formation of these volatile compounds e.g. high quantity of flavour compounds (pyrazines) were obtained at 167ºC for 22 minutes (Anese et al., 2011). Roasting conditions may vary depending on origin of the beans, the type of coffee beans (dry or washed) and moisture content of the coffee beans. For instance, at constant temperature, washed coffee beans may take longer time to roast compared to dry processed beans.

Fig.1.7: Schematic representation main volatile formation and reactions during the coffee roasting process (Ribeiro et al., 2009).

Chapter 1

1.9 Factors influence coffee quality and biochemical composition

The inherent quality and compositional characteristics of coffee beans are, however, controlled, among others, by interacting effects of genetic trait (Bertrand et al., 2006; Leroy et al., 2006 ), growing altitude (Avelino et al., 2005; Bertrand et al., 2006; Borém, 2014; Joët et al., 2014; Lara-Estrada & Vaast, 2007; Link et al., 2014; Vaast et al., 2005a; Villarreal et al., 2009), shade management (Avelino et al., 2007; Bosselmann et al., 2009; Bote & Struik, 2011; Lara-Estrada & Vaast, 2007; Muschler, 2001; Vaast et al., 2006), harvest periods (Guyot et al., 1988; Läderach et al., 2011; Vaast & Bertrand, 2005) and postharvest processing technique (Duarte et al., 2010; Joët et al., 2010; Selmar et al., 2002).

Genetic traits

The species and genotype of a plant play important role in the quality of the produced coffee (Ky et al., 2001; Leroy et al., 2006). Several studies have also shown that the genetic diversity of arabica coffee is low compared to that of robusta coffee (Cubry et al., 2007; Lashermes et al., 2011). As a result arabica is more susceptible to diseases and pests and give lower yield than robusta coffee. Large differences in chemical composition can be also found between these species resulting in major differences in beverage quality. For example, arabica coffee is characterized by its lower bitterness, lower caffeine and better flavour and aroma than robusta. Robusta is known as less aromatic and richer in caffeine than arabica coffee (Leroy et al., 2006).

A significant variation also exists within the species. The existence of wide diversity for quality trait of Ethiopian arabica coffee genotypes has been reported elsewhere (Abeyot et al., 2011; Kitila et al., 2011). Van der Vossen (1985), also reported significant differences in cup quality attributes among different arabica coffee cultivars and their crosses. Medium to high heritability was obtained for some cup quality attributes e.g. flavour, acidity and aroma (Abeyot et al., 2011). This shows that there is a huge potential resource for improvement coffee bean qualities via breeding. However, the impact of genotypes on quality differs based on of the ecological conditions where the plant grows. Hence, due attention should be given for genotype by environment interaction during breeding strategies to improve coffee quality.

General introduction

Growing altitude

The development of coffee bean quality conferring good quality with distinct characteristics depends on the environment the coffee is grown (Bote, 2016). Altitude is one of the environmental factors that affect coffee bean quality through optimizing temperature (Decazy et al., 2003; Lara-Estrada & Vaast, 2007; Vaast et al., 2005a). Temperature plays an important role in the phenologic cycle of coffee, particularly on fruit development and ripening (Guyot et al., 1996). The ideal temperature for arabica coffee is between 18-21ºC (Camargo et al., 1992) and coffee beans that are grown outside of these limits will have incomplete fruit maturation (Link et al., 2014). For instance, at lower altitude where temperature is beyond the optimum ranges, fruit ripening is accelerated leading to bean quality losses (Muschler, 2001; Vaast et al., 2006). When temperatures are high during the growing period, some biochemical compound e.g.butan-1, 3-diol, butan-2, 3-diol accumulate during the ripping and reduces the bean quality (Bertrand et al., 2012). Moreover, relatively high temperature during blossoming, especially if associated with a prolonged dry season, may cause flower abortion. Vaast et al. (2006) also reported that high temperature of low altitude decreases leaf lifespan resulting in an impaired leaf-to-fruit ratio leading to lower assimilate supply for developing beans.

It has been reported that altitude had a positive influence on the final beverage quality of the coffee (Avelino et al., 2005; Bertrand et al., 2006; Bosselmann et al., 2009). This is due to the highest altitude generate more optimal temperature for ideal cherry ripening that play an important role in development of attributes conferring superior quality (Silva et al., 2005; Vaast et al., 2006). Lower temperature prolongs the maturation period of coffee beans. This provides sufficient time for chemical biosynthesis (Joët et al., 2014; Silva et al., 2005; Vaast et al., 2006). Other similar studies also reported influences daily temperature on bean biochemical compounds (such as CGA composition) of coffee seeds at maturity (Joët et al., 2010). The study, suggested that temperature is a key environmental parameter that directly affects biosynthesis of coffee bean biochemical compositions.

The amount of chlorogenic acid, caffeine and sucrose increased with altitude pleasantly fruity character and intense flavor is obtained at higher than at lower altitude (Avelino et al., 2005; Avelino et al., 2007; Bertrand et al., 2006). However, in Bertrand et al. (2006),

Chapter 1 chlorogenic content varied across altitude ranges e.g. increased from 900m -1350 m but decreased from 1400-1450m. In some other studies e.g.Link et al. (2014), chlorogenic acid content reduced with altitude but the caffeine content increased. Sridevi and Parvatam (2014), on the other hand, reported the negative effect of altitude on caffeine content. This calls for further investigation to establish clear relationship between altitudes and these components. In addition, there is no any information on effect of very high altitude (e.g. greater than 1850 m) on both coffee quality and biochemical compositions of the beans.

Shade management

Coffee tree is grown mostly in association with trees that provide shade and other services (Van Oijen et al., 2010). Shade tree in coffee production system provides economic and ecological benefits (Yadessa et al., 2009). Some studies showed that coffee shade trees play significant roles in soil fertility management (Kimemia, 2007; Muleta et al., 2008), biodiversity conservation, carbon sequestration (Harmand et al., 2007; Perfecto et al., 2005) and income generation (Beer et al., 1998). It has also been reported that coffee shade agroforestry systems have gained increased attention as a cost-effective option for climate change mitigation (Denu et al., 2016; Field et al., 2014). Shade also reduce biannual bearing pattern and overbearing i.e. a tendency to produce large fruit loads in one year followed by low production in the next year due to competition for photoassimilates between coffee berries and developing young branch parts (Cannell, 1985; DaMatta, 2004). In addition shade creates more favorable microclimatic conditions for coffee growth via lowering temperature of up to 4°C under sub-optimal conditions of low altitude (< 700 m) and by up to 2°C under optimal conditions (> 1100 m) (Vaast et al., 2005). It has also been suggested that coffee trees grown under shade trees undergo slower and more uniform ripening. This allows the fruits to be developed in a more balanced manner giving good quality (Muschler, 2001).

In other studies, e.g.Vaast et al. (2006), shade level up to 40% positively affected bean composition and beverage quality by delaying berry flesh ripening by up to one month. In other studies e.g.Bosselmann et al. (2009) shade was, however, reported to have a negative effect on coffee quality attributes such as acidity, body and overall cup preferences. In contrast to this study,Avelino et al. (2007) observed differences on cup

General introduction quality and chemical compositions between coffees produce under open sun and shade trees. These results indicate that shade effects are site specific and calls a need for studying the relations between shade and cup quality attributes along environmental gradients (i.e. altitude).

Harvesting and postharvest processing

Harvesting

Besides the growing conditions, coffee bean quality depends on maturity and harvest periods. Coffee flowering is a complex sequence that is affected by climatic factors such as temperature. These complex sequences do not occur at the same time in every part of the plant, and result in different blossom and harvest periods. Coffee cherries can be harvested when fruit are red and fully mature. This improves coffee bean quality and marketability. As the cherries do not ripe all at the same time, several harvests are usually performed to collect beans from the mother trees. In Ethiopian, for example, coffee beans are harvested between October and February depending on growing conditions. The quality of these coffees may vary with different harvest periods. It has been reported that early harvested coffee beans had better quality than late harvested coffee beans (Bertrand et al., 2005; Läderach et al., 2011). Other Guyot et al. (1988) reported that late harvested coffee beans showed better beverage quality and larger bean size than early harvested beans. This also calls for further research to clearly ascertain effects of harvest periods on coffee bean quality.

Postharvest processing methods

The quality of coffee beans and its biochemical composition are not only depending on growing conditions and harvest periods, but also on postharvest processing methods. Some studies have shown that processing methods plays a significant role in determining chemical composition and final coffee beverage. The washed processed coffee beans have more aromatic with fruity and acidic attributes and possess lesser bitter, burnt and woody notes while dry processed coffee beans are generally characterized by more body (Duarte et al., 2010; Leloup et al., 2005; Selmar et al., 2002). Knopp et al. (2006), reported that dry and washed processing methods may also change

Chapter 1 the biochemical composition (e.g. free amino acids, organic acids) of the coffee beans. These underlined differences coffee quality attributes are possible due to the fermentation process involved in mucilage removal during washed processing. In the fermentation step, the mucilage (rich in polysaccharides) is hydrolyzed by enzymes from both the coffee tissues and from microorganisms (e.g. yeast and bacteria) found on the fruit skins (Belitz et al., 2009; Silva et al., 2013).This process leads to the production of alcohols and acids and other metabolic compounds that interfere in the final beverage coffee quality (Silva et al., 2013).

Numerous species of microorganisms have been isolated from the fermentation phase of washed processing. For instance, bacterial species: aerobic bacteria such as Klebsiella ozaenae, K. oxytoca, Erwinia herbicola and lactic acid bacteria such as Leuconostoc mesenteroides and Lactobacillus brevis. Yeast species such as Kloeckera apis apicualata, Candida guilliermondii, C. tropicalis, C. parapsilosis, Cryptococcus albidus, C. laurentii, Pichia kluyveri, Hanseniaspora uvarum, Schizosaccharomyces spp and Saccharomyces cerevisiae (Avallone et al., 2001; de Melo Pereira et al., 2014; Lee et al., 2015; Masoud et al., 2004). These microorganisms exhibit different metabolic processes and secrete different types of extracellular enzymes during fermentation which may lead to production of different aroma precursor’s compounds. This suggests that suitable species would modify the aroma precursors. For instance, Pichia fermentans and Yarrowia lipolytica were identified as suitable yeast species in enhancing the aroma attributes.

Although fermentation plays a great role in enhancing coffee aroma and flavour coffee undesirable fermentation process would have a negative effect on coffee aroma. Over- fermentation is one of the major problems observed among coffee producers and processors (Jackels et al., 2006; Puerta Quintero, 2001). This leads to the production of “black” or ‘‘stinker’’ beans. These beans are commonly associated with fruity, flora, sour and alcoholic attributes while under-fermentation (i.e. incomplete degradation of the mucilage layer) also promote the growth of undesirable microorganisms (Lee et al., 2015). During dry processing, microbial fermentation lead to the breakdown of the pulp and mucilage with the coffee which may occur during fruit drying (Evangelista et al., 2014; Schwan & Wheals, 2003). However, this fermentation process is significantly slower than during the washed processing (Evangelista et al., 2014).

General introduction

General, in dry processing the levels of maturity are not consistent among the harvested coffee cherries. For instance, in Ethiopian, all types of coffee cherries (unripe, ripe and overripe) are used. As the result washed processed coffee beans have superior quality and received higher premium price than dry processed beans. However, the majority of the coffee beans, 70 – 80% of coffee are dry processed while the remaining 20 – 30% are washed processed coffee (Bart et al., 2014).This may calls for further investigation to understand the influence dry and washed processing by using identical coffee cherries. Although several studies have been carried out to evaluate influence of altitude, shade, harvest periods and postharvest processing methods, there is still limited information on the interactive effect of these factors coffee quality and biochemical composition.

Mycotoxins

Recently, contamination by fungi is becoming a main constraint in coffee bean production. Temperature, moisture content and storage conditions are the major factors for mold development (Amézqueta et al., 2009). The presence of these fungi in coffee beans result in the production of mycotoxins. Mycotoxin is a secondary metabolite produced by fungus that grows on crops and a variety of feed- and food stuffs including coffee beans. It is capable of causing disease and death in both humans and animals. However, information on the presence of mycotoxins in Ethiopian coffee is not well documented. But recent study e.g. Geremew et al. (2016) assessed for the first time the presence of fungi and mycotoxins in coffee samples from the local market in Ethiopia in Jimma zone. From the result the incidence total toxigenic fungal incidence was found to be up to 87% which is mainly caused by Aspergillus (79%), Fusarium (8%) and Penicillium (5%). The incidence varied with coffee processing methods and storage conditions. But mycotoxins e.g. ochratoxin A (OTA) contamination on average found to be lower (1.5 mg kg-1) than the EU tolerable limit (5 mg kg-1) set for roasted coffees. In general further research is also needed to explore the presence of OTA for other for coffee growing regions for further interventions.

Chapter 1

Value chain actors, sorting and postharvest processing

In Ethiopia the majority of coffee export to international market originates from smallholder coffee farmers. Coffee beans produced by these smallholder farmers reach consumers by passing through different value chain actors. The actors at each stage add value to the final product. Cooperatives and private traders are among the value chain actors participating in Ethiopian coffee production, processing and marketing systems. Cooperatives and unions are recently introduced actors with the aim of improving the bargaining power of farmers for better prices and eventually to improve their livelihoods. For instance, coffee beans that pass through cooperative receive higher premium price ca. 16 % over the private traders (Bart et al., 2014).

In general the coffee quality that passes through the actors i.e. farmers to cooperatives or farmers to private traders are determined by the extent of the management applied by the actors during production, harvesting, processing, drying and storage. To produce good quality coffee beans, the entire value chain should be properly managed by responsible actors. For instance, at farm level during harvesting, proper coffee cherries e.g. ripe red cherry should be considered. This is mainly managed by coffee farmers. However, smallholder coffee farmers may not follow the standard procedures i.e. selective picking and also varies from farmers to farmers depending on the actors to whom the coffee cherries will be sold e.g. cooperatives or private traders. Moreover, coffee beans passes through cooperatives or private traders may vary due to the different growing environments and coffee cherry management e.g. sorting prior to processing and also due to management differences during processing methods. However, there were no scientific studies that assessed the impact of coffee farmers, cooperatives and private trader’s value chain on coffee bean quality. This, however, requires further studies to better understand the effectiveness of these value chains in relation to coffee cherry management at harvest (e.g. proper harvests) and after harvest i.e. sorting and postharvest processing methods.

Moreover, as the result of growing demand for healthier and eco-friendly produced coffee beans (Giovannucci et al., 2014; Jena et al., 2012; Stellmacher, 2007), coffee certification has also been introduced to the country. Certification became a new opportunity in the international market and intends to add value to a product and

General introduction consumers are also willing to pay extra for the certified product. Thus, certification attempts to give consumers a guarantee about the social or ecological sustainability of the mode of production of the products. Nowadays, a number of coffee cooperatives in Ethiopia are certified e.g. organic, fair trade, Rain Forest Alliance and UTZ and allow receiving premium price for their certified coffee beans (Bart et al., 2014; Jena et al., 2012).

Organic coffee refers to mode of production of coffee i.e. coffee grown without chemical fertilizer and pesticides where the biological value of the coffee harvested is equal to the value returned to the soil. This is achieved through practices such as mulching, application of compost and organic wastes and proper plantation of symbiotic shade trees. Fair trade coffee certifications are set by the Fairtrade Labelling Organizations (FLO). They guarantee a minimum price to smallholder farmers that promote greater economic incentives (Pierrot et al., 2010). Rainforest alliance certification is intended to integrate the biodiversity conservation, natural resource conservation and suitable agricultural practices to ensure sustainable farm management. It also makes provision for protecting the rights and welfare of workers and communities (ITC, 2011; Pierrot et al., 2010). UTZ certification promotes better business practices as a way to achieve sustainability. It also provides market transparency and traceability through promoting good agricultural practices at the farm level (Adam, 2009; Potts et al., 2014; Volkmann, 2008). Despite the increasing demand for certification and establishment of a “new” coffee value chain (cooperatives, unions) in Ethiopia, there are so far no studies that report impacts of cooperatives and certification on coffee bean quality. This call for further research to ascertain the effect of key actors in the value chain: farmers, certified and non-certified cooperatives and private traders. In general, to the best of our knowledge, coffee quality has not been yet much investigated in Ethiopia. Hence, this requires investigation and analysis of factors that affect coffee quality along the value chain.

Chapter 1

1.10 Objectives and outline of the PhD thesis

The general aim of the thesis was to explore the influence of growing environments, management techniques and value chain actors on quality and biochemical composition of Arabica coffee beans. This will contribute to derive concepts and methodologies to enhance coffee quality in Ethiopia.

The specific objectives are to:

¾ Investigate the interactive effects of altitude, shade and harvest period on physical characteristics, cup quality attributes and selected biochemical composition. ¾ Investigate the physical quality, cup quality attributes and biochemical composition of coffee beans in relation to different growing altitudes, postharvest processing methods and their possible interactions. ¾ Examine physical quality and cup quality attributes of coffee beans that pass through certified cooperatives, non-certified cooperatives and private traders. ¾ Assess the variability of coffee quality among different coffee farmers who are a member cooperatives (certified and non-certified cooperatives) and non- member cooperatives. ¾ Examine the impact of coffee cherry sorting and postharvest processing methods on physical and cup quality attributes. ¾ Develop NIRS-based model to predict coffee cup quality from green coffee beans spectra.

For the study semi-forest coffee production system in Jimma zone, Ethiopia was considered. Fig.1.8 indicates overview and structure of the thesis. The thesis has six chapters. Chapter 1 gives the general introduction. Chapters 2 to 5 present the main findings of the study. Throughout these chapters coffee bean quality is described by physical quality, preliminary cup quality, preliminary total quality, total specialty cup quality and its specific quality attributes (e.g. overall cup preferences, acidity, body, aroma, flavor and aftertaste) and biochemical composition (caffeine, total chlorogenic acid and sucrose). Chapter 2 addresses the interactive effects of growing altitude: high (1950–2100 m asl.), mid (1600-1680 m asl.), shade: open (0%), medium (40-55%) and dense (65-85%) and harvest periods (early, middle and late) on coffee quality and biochemical composition. Chapter 3 describes the interactive effect of growing altitude: 32

General introduction high (1950 – 2100 m asl.), mid (1600 – 1650 m asl.), and low (1450-1550 m asl.) and postharvest processing methods (dry, washed and semi-washed methods) on coffee quality and biochemical composition. Chapter 4 explains how different coffee actors: farmers, cooperatives, certification and private actors affect coffee bean quality. In this chapter coffee sorting and postharvest processing method (dry and washed methods) on coffee quality will also be shown. In chapter 5, the description of the model developed from green bean NIR spectra and its validation will be elaborated and discussed. In chapter 6 of the thesis, summary of the findings indicted in the previous chapters (2 to 5) is presented. This chapter will also contain general conclusions and prospective for future research.

Fig.1.8: Outlines of the PhD thesis

Chapter 2

Influence of growing altitude, shade and harvest period on quality and biochemical composition of Ethiopian specialty coffee

This chapter has been published as

Tolessa, K., D'heer, J., Duchateau, L and Boeckx, P. (2016). Influence of growing altitude, shade and harvest period on quality and biochemical composition of Ethiopian specialty coffee. Journal of the Science of Food and Agriculture, in press DOI: 10.1002/jsfa.8114.

Chapter 2

Abstract

Coffee quality is key characteristic for the international market, cup quality and chemical bean constituents describe it. In Ethiopia, using total specialty cup scores, coffees are grouped into Q1 (specialty 1) ≥ 85 and Q2 (80 - < 85). This classification results in market segmentation and higher prices. Although different studies evaluated effects of altitude and shade on bean quality, optimum shade levels along different altitudinal ranges are not clearly indicated. Information on effects of harvest periods on coffee quality is also scanty. This study examined influences of these factors and their interactions on Ethiopian coffee quality. Coffee from high altitude with open or medium shade and early to middle harvest periods gave superior bean quality. These growing conditions also favoured production of beans with lower caffeine. Increasing altitude from mid to high, ca. 400m decreased caffeine content by 10%. At high altitude, dense shade decreased Q1 coffee by 50%. Compared to late harvesting, early harvesting increased the percentage from 27 to 73%. At mid altitude >80% is Q2 coffee. Changes of quality scores driven by altitude, shade and harvest period are small, though may induce dramatic switches in the fraction Q1 vs. Q2 coffee. The latter affects farmers’ profit and competitiveness in the international markets.

Key words: Arabica coffee; altitude; shade; cup quality; caffeine; chlorogenic acid

Influence of growing altitude, shade and harvest period

2.1 Introduction

Coffee is the world's most traded commodity after oil (Pendergrast, 2009) and its quality is a key characteristic for the international coffee market (Jeszka-Skowron et al., 2015; Lara-Estrada & Vaast, 2007). Coffee quality is determined by organoleptic characteristics (cup quality), physical appearance and chemical constituents (Agwanda et al., 2003; Joët et al., 2010). The increase in demand and consumption of high quality, single origin, specialty coffee with specific characteristics results in market segmentation (e.g. specialty coffee market) and creates a strong potential and new opportunities for coffee- producing countries (Läderach et al., 2011; Oberthür et al., 2012; Wood, 2015). Market segmentation for specialty coffee attracts new customers and results in higher coffee prices (Labouisse et al., 2008). For instance, specialty coffee beans receive a premium price that is ca. 20-50% higher compared to regular coffee beans (Bart et al., 2014).

Ethiopia is one of the top ten coffee-producing countries in the world and the largest exporter in Africa (Bart et al., 2014). The country is naturally gifted with a suitable climate and has the potential to produce single origin specialty Arabica coffee beans with a wide range of flavors (Coste et al., 1992; ICCO, 2014). Recently, nine single origin specialty coffees (Jimma, Nekemte, Illubabor, Limu, Tepi, Bebeka, Yirgacheffe, Sidamo and Harar) were identified and entered into trade circuits of the world coffee market. Among them, the sundried coffee beans from ‘Harar’, the so-called ”Mocca” and the washed beans from ‘Yirgacheffe’ are considered as the finest and best quality coffee (MFA, 2016).

Currently, specialty coffee accounts for ca. 20% of Ethiopia’s coffee export and there is also a very high potential to boost its share in the world market (Robert, 2015). Thus, an improved specialty coffee market is thought to be beneficial for Ethiopia to remain competitive in the international market. Ethiopian producers will benefit from high quality coffee if its supply remains stable (Oberthür et al., 2012) and if the global coffee chain is not changing as a result of deregulation, new consumption patterns and/or evolving corporate strategies (e.g. branding) (Ponte, 2002). Producers, on the other hand, should implement improved, sustainable agronomic practices (e.g. shade-grown coffee) to produce beans desired by consumers. Along the coffee supply chain, however, numerous factors affecting coffee quality have been identified. Genetic traits (Leroy et al., 2006), growing environment (Avelino et al., 2005; Decazy et al., 2003; Muschler, 2001;

Chapter 2

Vaast et al., 2005) and postharvest processing methods (Joët et al., 2010; Selmar et al., 2006) are known to predominantly affect coffee quality. The most important environmental factor, most commonly linked to influence coffee quality, is the altitude where coffee is grown (Avelino et al., 2005; Muschler, 2001). Some authors reported that the best quality of Arabica coffee comes from higher altitude as a result of lower daily temperatures, providing a slower ripening of the beans and allowing more time for bean filling.

Shade is also found to favour a sustainable production of high coffee quality, especially under suboptimal conditions where temperatures are higher than optimal (Muschler, 2001; Vaast et al., 2005). Some authors indicated that shade reduces temperature stress in the canopy and lengthens the maturation period of coffee berries. It also reduces periodic over-bearing and a subsequent die back of coffee plants (Alemu, 2015). In line with this, Läderach et al. (2011) reported that coffee quality scores increase with the level of shading. Bosselmann et al. (2009), on the other hand, reported that at high altitude shade had no significant effect on cup quality attributes. At lower altitude, however, shade reduced the number of small beans with no significant effect on coffee sensorial attributes. In the study by Lara-Estrada and Vaast (2007), shade was reported not to have any significant influence on coffee organoleptic properties in any of the conditions of altitude and fertilization. These contradicting reports show that optimum shade management practices are highly site specific and need further study to clearly indicate its influence on coffee quality attributes across different altitudinal gradients.

In Ethiopia, coffee trees are grown under shade of different levels and shade management is among the dominant agronomic practices in traditional organic coffee growing systems (Alemu, 2015; Yadessa et al., 2009). Recent studies showed that shade significantly affected yield (Kufa et al., 2007) and cup quality attributes of Ethiopian coffee (Bote & Struik, 2011; Yadessa et al., 2009). Although different studies have been carried out to evaluate the effect of shade on coffee yield and quality, optimum shade levels along different environmental gradients were not yet investigated in the country. In addition, there was no any study carried out on how altitude and shade interactively affect cup quality and biochemical composition of Ethiopian coffee beans.

In addition to shade and growing altitude, coffee harvesting periods were also reported to affect coffee quality. Guyot et al. (1996) indicated that late harvested coffee beans had

Influence of growing altitude, shade and harvest period

better beverage quality and larger bean size than early harvested beans. Bertrand et al. (2005), on the other hand, reported that early harvested coffee beans were better in beverage quality than late harvested beans. In another study, early harvesting was found to improve coffee quality compared to late harvesting (Läderach et al., 2011). This also calls for further research to clearly ascertain effects of harvest periods on coffee bean quality. To the best of our knowledge, there exist no studies analysing the interactive effects of growing altitude, shade levels and harvest periods on Ethiopian coffee bean quality. Hence, the main objective of this study was to examine their interactive effects on physical characteristics, cup quality attributes and selected biochemical compounds.

2.2 Materials and Methods

2.2.1 Study site

The study was carried out from October 2013 to Feburary 2014 in the Mana district, Jimma zone, Oromia regional state, south western Ethiopia. Mana is known as one of the predominant Coffea arabica L. growing areas. It is located at altitudinal ranges between 1470 to 2610 m asl, at 8o 67' N and 37o 07' E. The district has a total area of 47,898 ha of which 23% is at low (<1550 m asl; metres above sea level), 65% at mid (1550-1750 m asl) and the remaining 12% at high altitude (>1750 m asl) (JARC, 1996). The annual average temperature and rainfall are 20.5°C and 1523 mm, respectively. The soil type in the study area is uniform and described as Nitosol, with a pH ranging from 4.5 to 5.5 (ARDO, 2008).

2.2.2 Treatments and experimental design

Two altitudinal ranges were selected: high (1950 – 2100 m asl) and mid altitude (1600 – 1680 m asl) with an average temperature of thirty years of 19.3±1.9°C and 22.5±2.6°C, respectively. Different shade levels of ten coffee farms (five from each altitudinal range) with comparable coffee ages (7 to 10 years) were selected taking care that agronomic management practices did not differ substantially among the farms. Shade levels above canopies of the coffee trees in each farm were quantified using the Sunscan canopy analysis system (Type SS1, BF5-RPDA1, Delta-T Devices, Cambridge, UK) and grouped into three levels: open (no shade), medium (40-55%) and dense shade (65-85%). During measurements, the reference sensor was placed in open sun (no shade) and the shade

Chapter 2

percentage was determined in relation to this reference point. The shade percentage was quantified based on the photosynthetically active radiation (PAR) level of shade treatment in relation to PAR level of the open sun conditions. This has been done twice (i.e. during the experimental set up and at bean filling stage). The selected coffee trees were those trees found in shade tree relatively having uniform distribution however, the shade tree diversity of the farms were not included as treatment since the study mainly focused on shade percentage which are given by the shade tree. However, at high altitude the dominant shade trees are Acacia where as at mid altitude, Croton and Albizia are the dominant tree species

In the harvest season, from October 2013 to February 2014, red ripe cherries from each altitude and shade level were harvested at three different periods (early, middle and late). Early and middle harvesting at mid altitude of open and medium shades were carried out from the first to the third week of October 2013; while late harvested beans were collected from the first to the third week of November, 2013. Coffee samples at high altitude with open to medium shade levels were harvested early (first week of December 2013), middle (end of December 2013) and late (third week of January 2014). Under dense shade, on the other hand, early harvest was done in the fourth week of December 2013, middle harvest in the third week of January 2014 and late harvest in the first week of February, 2014. Split-split plot design was used for the experiment because of the complexity of the treatment e.g. altitude cannot be randomized and treated as whole plot with two levels (mid and high), shade as sub-plot with three levels (open, medium and dense) and harvest periods as sub-sub-plot with three levels (early, middle and late) with five replications (farms) at each attitude

2.2.3 Coffee quality analysis

Coffee cherries were sundried immediately after harvest on raised beds according to standard agronomic practices until a moisture content of ca. 11.5%. The dried coffee cherries were then dehusked using a coffee hulling machine (Coffee huller, McKinnon, Scotland) at Jimma University College of Agriculture and Veterinary Medicine (JUCAVM) and stored at room temperature being packed in hermetic plastic bags. 100-bean weight (g) of the dried beans was determined in three replicates using a digital sensitive balance (CTG-6H+, Citizen, Piscataway, NJ USA) and the bean diameter (mm) was measured on

Influence of growing altitude, shade and harvest period

45 randomly selected beans per treatment, using a digital calliper (Mitutoyo, IP 67, CD- 20-PPX, Kawasaki, Japan).

For quality analysis, 350 g of each coffee sample was sent to the Ethiopia Commodity Exchange (ECX), Jimma Branch, for physical and cup quality analysis, out of which 100 g was roasted at 200°C for 8-12 min using a roasting machine (Probat, 4 Barrel Roaster, Germany) adjusted to medium roasting. The roasting time is varied which depend on the moisture content of the beans and the types of coffee beans (dry or washed processed beans). Then the roasters are followed the colour change during roasting until it reach medium roasting which determine optimum roasting time for a given coffee sample. The medium roasting measured based on the M-basic (gourment) Agtron scale of approximately 58 on whole beans. Generally, roasting has been carried out based on the standard protocol developed by ECX for the commercial coffee quality analysis and grading purpose (John, 2011).

The roasted beans were tipped out into a cooling tray and rapidly cooled by blowing cold air through the beans for four min and then ground with a coffee grinding machine (Mahlkonig, K32SB2, Germany). 13.75 g of ground coffee was diluted in 250 mL hot water (93°C) to prepare an infusion. Five cups of brewed coffee of each coffee sample were prepared for analysis and a team of three professional cuppers, who operate in ECX, tasted and gave a score for each of the five cups.

The preliminary total quality score comprises physical (40/100) and preliminary cup quality (60/100) scores. The criteria commonly used to evaluate the physical quality of coffee beans include the presence of defects: primary (e.g. full black and sour beans) and secondary (e.g. partial black, insect damaged and broken beans) defects and odour (Santos et al., 2012). Cup quality was evaluated based on acidity, body, flavour and cup cleanness scores. Each of these cup quality attributes has a score between 0 and 15. The sum of these four cup quality attributes gives preliminary cup quality with a score between 0 and 60. Hence, the sum of physical and preliminary cup quality ranges between 0 and 100; the total sum resulted in ‘preliminary total quality’ scores. Preliminary total quality scores were used to classify the samples into different grades (grade 1= 91- 100, grade 2 = 81 - 90, grade 3 = 71 - 80 and grade 4 = 63 - 70) (see detail in chapter 1; Table 1.1). The scores of all samples were higher than 70, cup quality for specialty coffees (coffees with grades ranging from 1 to 3) and further speciality cup quality

Chapter 2

analysis was done based on overall cup preference, acidity, body, aroma, flavour, aftertaste, uniformity, cup cleanness, sweetness and balance, each on a scale ranging from 0 to 10. Based on the total specialty scores, coffee samples were further grouped into specialty 1 (Q1) ≥ 85, specialty 2 (Q2) (80-84.75) and a regular commercial coffee (< 80) (John, 2011). The grading has been carried out using an old ECX guideline (see details in chapter 1; Table 1.2). However, recently, ECX established a new grading system that reduced the old ten preliminary total quality levels to five levels: grade 1 ≥ 85, grade 2 = 75-84, grade 3 = 63-74, grade 4 = 47-62 and grade 5 = 31-46. (Lemma, 2015) Coffee samples that fall in grade 1 and 2 qualify for specialty coffee and are further grouped into specialty 1 (Q1) ≥ 85 and specialty 2 (Q2) (80 - 84.75). The preliminary total quality assessment scores for grade 2 should be higher than 80 to be qualified as specialty 1 (Q1).

2.2.4 Caffeine and chlorogenic acid extraction

Green coffee beans were ground (<0.5 mm) prior to extraction using a grinder (IKA, M20, Germany). The extraction was done using 100 mg ground coffee beans with 10 mL of methanol:water:acetic acid (30:67.5:2.5, v:v:v) containing 2 mg mL-1 ascorbic acid, according to Alonso-Salces et al. (2009). The extract was placed in an ultrasonic bath for 15 min and filtered over a 0.45μm PTFE filter prior to injection into a liquid chromatography (LC) system.

2.2.4.1 Liquid chromatography analyses

Using an established liquid chromatography - time of flight - mass spectrometry (LC- TOF-MS) method (Alonso-Salces et al., 2009), four subclasses of chlorogenic acids were identified in the green coffee samples: caffeoylquinic acid (CQA), feruloylquinic acid (FQA), dicaffeoylquinic acid (diCQA) and feruloyl-caffeoylquinic acid (FCQA). The quantification was done using LC equipped with a photodiode array (PDA) detector (Surveyor, Thermo Finnigan, USA) at 280 nm for caffeine and 320 nm for chlorogenic acids. The chromatographic separation was done using a prevail C18 (250 x 4.6 mm, 5 μm, Alltech, Belgium) column, kept at 25°C. The chromatographic method was based on Alonso-Salces et al. (2009) using a combination of 0.2% acetic acid in water (v/v) and HPLC grade methanol. The injection volume was 50 μL and a flow rate 1 mL min-1. Calibration was done by injecting concentration series of caffeine and chlorogenic acid

Influence of growing altitude, shade and harvest period

every 30 samples, using an adapted response factor for the other chlorogenic acid types. Total chlorogenic acid concentration is reported as the sum of all above individual chlorogenic acids. Data acquisition and processing were done using ChromQuest 4.1 SP2 software.

2.2.5 Moisture content determination

To express the data on dry weight basis instead of on wet matter basis, the moisture content of the coffee samples was determined according to AOAC procedures (AOAC, 2000).

2.2.6 Data analysis

Data were analysed with a Statistical Analysis System software (v. 9.2, SAS Institute Inc., Cary, NC USA,) using the mixed model procedure for a split-split plot design with altitude as the main plot, shade as sub-plot and harvest period as sub-sub plot. Significant difference between treatments means were separated using Tukey’s honest significant difference, the HSD test.

2.3 Results

2.3.1 Coffee quality

Significant interactions between altitude and shade were observed for total specialty coffee cup quality and its specific specialty cup quality attributes (overall cup preference, acidity, body, flavour and aftertaste) (P < 0.01; Table 2.1). Quality scores for total specialty cup quality (86.5), overall cup preference (8.2), acidity (8.3), body (8.2), flavour (8.0) and aftertaste (8.0) were higher for coffee beans grown at high altitude combined with open or medium shade level. The interaction between altitude and harvest period significantly influenced physical quality, overall cup preferences and acidity of coffee beans (Table 2.2). Coffee beans grown at mid altitude and collected during late harvest period had the lowest physical quality score (34.0), while scores for overall cup preferences and acidity were higher for coffee beans grown at high altitude and harvested in early or middle harvest periods (Table 2.2). Harvest period by shade

Chapter 2

interactions were not significant for any of the above variables. There were also no three- way interactions between altitude, shade and harvest period on any of the coffee bean quality attributes.

Altitude as a main effect (Table 2.3) indicated that preliminary cup quality, preliminary total quality, total specialty cup quality and aroma were lower at mid altitude. The scores for each of these quality attributes increased with increasing altitude (Table 2.3). Coffee beans grown in dense shade had a higher physical quality score (36.7). Coffee grown at open or medium shade gave a higher total specialty cup score (84.5) and overall cup preference (7.9). Beans harvested at early and middle harvest period were generally higher in preliminary cup quality, preliminary total quality, total specialty cup quality, overall cup preference and body scores compared to late harvested beans (Table 2.3).

Influence of growing altitude, shade and harvest period

Table 2.1: Interactive effect of altitude (A): high (1950–2100 m asl), mid (1600-1680 m asl) and shade (S): open (0%), medium (40-55%) and dense (65-85%) on physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty cup quality scores A S PQ PCQ PTQ TSCQ Specific specialty cup quality attributes OCP 1 Acidity Body Aroma Flavour AT 2 High Open 35.6 ± 0.4a 49.6 ± 0.8a 85.2 ± 0.8a 86.5 ± 0.9a 8.2 ± 0.2a 8.3 ± 0.1a 8.2 ± 0.2a 8.1 ± 0.2ab 8.0 ± 0.2a 8.0 ± 0.2a Medium 36.6 ± 0.4a 49.6 ± 0.7a 86.2 ± 0.8a 86.5 ± 0.8a 8.3 ± 0.1a 8.3 ± 0.1a 8.1 ± 0.1ab 8.2 ± 0.1a 7.9 ± 0.1a 8.0 ± 0.2a a a a b b b bc bc b b Dense 36.8 ± 0.6 47.6 ± 0.8 84.4 ± 0.8 83.7 ± 0.7 7.6 ± 0.1 7.9 ± 0.2 7.7 ± 0.1 7.7 ± 0.1 7.6 ± 0.1 7.5 ± 0.1 Mid Open 35.4 ± 0.4a 45.6 ± 0.4a 81.0 ± 0.7a 82.6 ± 0.3b 7.6 ± 0.1b 7.6 ± 0.1bc 7.6 ± 0.1c 7.5 ± < 0.1c 7.5 ± 0.1b 7.4 ± 0.1b Medium 36.2 ± 0.4a 46.6 ± 0.4a 82.8 ± 0.6a 83.2 ± 0.3b 7.6 ± 0.1b 7.8 ± < 0.1bc 7.7 ± 0.1bc 7.6 ± 0.1c 7.5 ± 0.1b 7.5 ± <0.1b Dense 36.6 ± 0.5a 45.6 ± 0.4a 82.2 ± 0.7a 83.4 ± 0.2b 7.6 ± <0.1b 7.8 ± 0.1ab 7.7 ± 0.1bc 7.6 ± 0.1bc 7.6 ± <0.1b 7.5 ± 0.1b

P-value 0.96 0.22 0.29 0.006 <0.0001 0.006 0.008 0.023 0.007 0.0025 1 Overall cup preference, 2 Aftertaste, different letters in the same column indicate significant difference according to tukey’s HSD post-hoc test (P<0.01), results are shown as mean ± standard error.

Table 2.2: Interactive effect of altitude (A): high (1950 –2100 m asl), mid (1600-1680 m asl) and harvest periods (HP): early, middle and late on the physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty coffee cup quality scores A HP PQ PCQ PTQ TSCQ Specific specialty cup quality attributes OCP1 Acidity Body Aroma Flavour AT2 High Early 36.8 ± 0.5a 49.8 ± 0.7a 85.6 ± 0.8a 86.9 ± 0.8a 8.3 ± 0.2a 8.5 ± 0.1a 8.3 ± 0.1a 8.1 ± 0.1a 8.1 ± 0.2a 8.1 ± 0.2a Middle 36.6 ± 0.6a 49.2 ± 0.6a 85.8 ± 0.6a 85.4 ± 0.8a 8.1 ± 0.2a 8.2 ± 0.1ab 7.9 ± 0.1a 8.0 ± 0.2a 7.8 ± 0.1bc 7.8 ± 0.1a Late 34.0 ± 0.4b 47.8 ± 0.9a 84.4 ± 0.9a 84.3 ± 0.8a 7.7 ± 0.1b 7.9 ± 0.1b 7.8 ± 0.1a 7.9 ± 0.1a 7.7 ± 0.1bc 7.8 ± 0.2a Mid Early 36.6 ± 0.5a 46.8 ± 0.4a 83.4 ± 0.7a 83.1 ± 0.3a 7.6 ± 0.1b 7.7 ± 0.1b 7.6 ± 0.1a 7.6 ± <0.1a 7.5 ± 0.1c 7.5 ± 0.1a Middle 36.6 ± 0.2a 46.4 ± 0.4a 83.2 ± 0.4a 83.3 ± 0.4a 7.7 ± 0.2b 7.8 ± 0.1b 7.7 ± 0.1a 7.5 ± 0.1a 7.5 ± 0.1c 7.4 ± 0.1a Late 34.0 ± 0.4b 44. 6± 0.3a 79.4 ± 0.4a 82.7 ± 0.2a 7.6 ± 0.1b 7.7 ± <0.1b 7.5 ± <0.1a 7.5 ± <0.1a 7.5 ± <0.1c 7.5 ± 0.1a P-value 0.004 0.93 0.067 0.29 0.007 0.0008 0.09 0.86 0.012 0.35 1 Overall cup preference, 2 Aftertaste, different letters in the same column indicate significant difference according to Tukey’s, HSD post hoc test (P<0.01); results are shown as mean ± standard error.

Chapter 2

Table 2.3: Physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty cup quality scores of coffee beans as affected by altitude (A): high (1950–2100 m asl), mid (1600-1680 m asl), shade (S): open (0%), medium (40-55%) and dense (65-85%) and harvest periods (HP): early, middle and late Factor PQ PCQ PTQ TSCQ Specific specialty cup quality attributes OCP1 Acidity Body Aroma Flavour AT2 A High 36.3 ± 0.3a 48.9 ± 0.5a 85.2 ± 0.5a 85.5 ± 0.5a 8.0 ± 0.1a 8.2 ± 0.1a 7.9 ± 0.1a 7.9 ± 0.1a 7.9 ± 0.1a 7.9 ± 0.1a Mid 36.1 ± 0.3a 45.9 ± 0.3b 82.6 ± 0.4b 83.0 ± 0.2b 7.6 ±< 0.1b 7.7 ± 0.1b 7.6 ± <0.1b 7.5 ±<0.1b 7.5 ±<0.1b 7.5 ± 0.1b P-value 0.57 <0.0016 <0.0025 <0.004 <0.024 <0.012 0.011 <0.008 0.03 0.016 S Open 35.5 ± 0.3b 47.6 ± 0.6ab 83.3 ± 0.7a 84.5 ± 0.6ab 7.9 ± 0.1a 7.9 ± 0.1a 7.9 ± 0.1a 7.8 ± 0.1a 7.7 ± 0.1a 7.7 ± 0.1ab Medium 36.4 ± 0.3ab 48.1 ± 0.5a 84.5 ± 0.6a 85.7 ± 0.5a 7.9 ± 0.1a 8.0 ± 0.1a 7.9 ± 0.1a 7.8 ± 0.1a 7.7 ± 0.1a 7.8 ± 0.1a Dense 36.7 ± 0.4a 46.6 ± 0.5b 83.1 ± 0.6a 83.6 ± 0.4b 7.6 ± 0.1b 7.9 ± 0.1a 7.7 ± 0.1a 7.7 ± 0.1a 7.6 ± 0.1a 7.5 ± 0.1b P-value 0.008 0.033 0.06 0.009 0.0003 0.37 0.055 0.148 0.17 0.04 HP Early 36.2 ± 0.4ab 48.3± 0.5a 84.5 ± 0.6a 84.9 ± 0.6a 7.9 ± 0.1a 8.1 ± 0.1a 7.9 ± 0.1a 7.8 ± 0.1a 7.8 ± 0.1a 7.8 ± 0.1a Middle 36.7 ± 0.3a 47.8 ± 0.5a 84.5 ± 0.4a 84.4 ± 0.5a 7. 9± 0.1a 7.9 ± 0.1ab 7.8 ± 0.1ab 7.8 ± 0.1a 7.7 ± 0.1ab 7.6 ± 0.1a Late 35.7 ± 0.3b 46.2 ± 0.6b 81.9 ± 0.7b 83.5 ± 0.4b 7.6 ± 0.1b 7.8 ± 0.1b 7.8 ± 0.1b 7.7 ± 0.1a 7.6 ± 0.1b 7.6 ± 0.1a P-value 0.06 0.004 0.0009 0.003 0.006 0.005 0.027 0.37 0.04 0.32 1 Overall cup preference, 2 Aftertaste, different letters in the same column indicate significant difference according to Tukey’s, HSD post hoc test (P<0.01); results are shown as mean ± standard error.

Influence of growing altitude, shade and harvest period

2.3.2 100-bean weight, caffeine and total chlorogenic acid content

Three way interactions between altitude, shade and harvest periods were significant for caffeine content (P < 0.01); Table 2.4). Highest caffeine content (17.9 g kg-1) was obtained from mid altitude with dense shade and early harvested beans; while the lowest content (14.5 g kg-1) was from high altitude with medium shade and middle harvested beans (Table 2.4). No three way interactions were found either for 100-bean weight or for total chlorogenic acids contents of coffee beans.

Altitudes by shade interactions were significant for total chlorogenic acid content (P < 0.01), but not for 100-bean weight and caffeine contents. Total chlorogenic acid contents decreased from mid to high altitude and from dense to open shading at high altitude (Table 2.5). Coffee beans from mid altitude and open sun had a higher total chlorogenic acid content (46.5 g kg-1), while beans from high altitude and open shading gave 40.5 g kg-1. The interactions between altitude and harvest period were not significant for any of the 100-bean weight, caffeine or total chlorogenic acid content. The interactions between shade and harvest periods had a significant effect on 100-bean weight, but not on caffeine and chlorogenic acid content (Table 2.6). Coffee beans produced in dense or medium shade gave higher bean weights at any harvest period.

Table 2.4: Interactive effect of altitude (A): high (1950–2100 m.asl), mid (1600-1680 m.asl), shade (S): open (0%), medium (40-55%) and dense (65-85%) and harvest periods (HP): early, middle and late on caffeine content Altitude Shade Harvest period Caffeine (g kg -1) High Open Early 15.9 ± 0.2de Middle 15.1 ± 0.3ef late 15.1 ± 0.4ef Medium Early 17.1 ± 0.2abc Middle 14.5 ± 0.3f late 15.1 ± 0.5ef Dense Early 15.0 ± 0.3ef Middle 16.4 ± 0.4cd late 15.2 ± 0.3fe Mid Open Early 16.7 ± 0.3bcd Middle 17.2 ± 0.3abc late 16.4 ± 0.3cd Medium Early 17.1 ± 0.5abc Middle 17.8 ± 0.4a late 16.9 ± 0.3a-d Dense Early 17.9 ±0.1a Middle 17.7 ±0.5b late 17.1 ± 0.4abc P-value 0.0002 Different letters in the same column indicate significant difference according to Tukey’s HSD post hoc test (P<0.01); results are shown as mean ± standard error; expressed as g kg-1 in dry weight basis.

Chapter 2

The main effects are given in Table 2.7. Altitude affected both the caffeine content and 100-bean weight. Coffee beans produced at mid altitude had a higher caffeine content (17.2 g kg-1) than beans produced at high altitude (15.5 g kg-1), while the bean weight was 16.9 and 14.1 g at high and mid altitude, respectively. Beans grown at medium or dense shade were ca. 10% heavier than beans in open sun. The harvest period as main factor, on the other hand, had no effect on 100-bean weight, caffeine and total chlorogenic acid contents.

Table 2.5: Interactive effect of altitude (A): high (1950–2100 m asl), mid (1600-1680 m asl) and shade (S): open (0%), medium (40-55%) and dense (65-85%) on 100-bean weight, caffeine and total chlorogenic acid content A S 100-bean Caffeine Total chlorogenic -1 -1 weight (g) (g kg ) acid (g kg ) a a c High Open 16.1 ± 0.3 15.4 ± 0.2 40.5 ± 0.9 a a bc Medium 17.5 ± 0.3 15.6 ± 0.4 42.7 ± 0.8 a a bc Dense 17.0 ± 0.2 15.6 ± 0.2 43.2 ± 0.7 a a a Mid Open 13.3 ± 0.2 16.8 ± 0.2 46.5 ± 0.8 a a b Medium 14.2 ± 0.2 17.3 ± 0.2 43.7 ± 0.86 a a ab Dense 14.7 ± 0.2 17.6 ± 0.2 45.4 ± 1.0 P-value 0.16 0.34 0.0046 Different letters in the same column indicate significant difference according to Tukey’s HSD post hoc test (P<0.01); results are shown as mean ± standard error; expressed as g kg-1 in dry weight basis for caffeine and total chlorogenic acid.

Table 2.6: Interactive effect of shade (S): open (0%), medium (40-55%) and dense (65-85%) and harvest periods (HP): early, middle and late on 100-bean weight, caffeine and total chlorogenic acid content S HP 100-bean Caffeine Total chlorogenic -1 -1 weight (g) (g kg ) acid (g kg ) Open Early 13.7 ± 0.3bc 16.3 ± 0.2d 44.9 ± 1.2a Middle 15.2 ± 0.3b 16.2 ± 0.4cd 43.1 ± 1.5a Late 15.2 ± 0.6b 15.7 ± 0.3bcd 42.4 ± 1.4a Medium Early 15.9 ± 0.5a 17.2 ± 0.3a 44.3 ± 1.0a Middle 15.8 ± 0.5a 16.1 ± 0.6d 42.3 ± 0.9a Late 15.8 ± 0.a 15.9 ± 0.4cd 42.9 ± 0.8a Dense Early 15.8 ± 0.4a 16.5 ± 0.5abc 45.5 ± 1.2a Middle 15.8 ± 0.5a 17.0 ± 0.4ab 42.1 ± 1.1a Late 15.9 ± 0.5a 16.1 ± 0.4cd 45.1 ± 0.7a P-value 0.009 0.049 0.35 Different letters in the same column indicate significant difference according to Tukey’s HSD post hoc test (P<0.01); results are shown as mean ± standard error; expressed as g kg-1 in dry weight basis for caffeine and total chlorogenic acid.

Influence of growing altitude, shade and harvest period

Table 2.7: 100-bean weight, caffeine and total chlorogenic acid as affected by altitude (A): high (1950– 2100 m asl) and mid (1600-1680 m .asl), shade (S): open (0%), medium (40-55%) and dense (65-85%) and harvest periods (HP): early, middle and late Factor Level 100-bean Caffeine Total chlorogenic -1 -1 weight (g) (g kg ) acid (g kg ) A High 16.9 ± 0.2a 15.5 ± 0.2b 42.1 ± 0.5b Mid 14.1 ± 0.2b 17.2 ± 0.1a 45.2 ± 0.5a P-value 0.0009 <0.0001 0.012 S Open 14.7 ± 0.3b 16.1 ± 0.2a 43.5 ± 0.8a Medium 15.9 ± 0.3a 16.4 ± 0.3a 43.2 ± 0.6a Dense 15.9 ± 0.3a 16.6 ± 0.3a 44.3 ± 0.7a P-value <0.0001 0.059 0.311 HP Early 15.2 ± 0.3a 16.7 ± 0.2a 44.9 ± 0.7a Middle 15.6 ± 0.3a 16.5 ± 0.3ab 42.5 ± 0.8b Late 15.6 ± 0.4a 15.9 ± 0.2b 43.5 ± 0.6ab P-value 0.27 0.015 0.039 Different letters in the same column indicate significant difference according to Tukey’s HSD post hoc test (P<0.01); results are shown as mean ± standard error; expressed as g kg-1 in dry weight basis caffeine and total chlorogenic acid.

2.4 Discussion

Several studies reported influences of altitude (Avelino et al., 2005; Guyot et al., 1996) and shade (Bote & Struik, 2011; Geromel et al., 2006; Muschler, 2001; Vaast et al., 2006) on coffee bean quality. The present study, however, expanded the scope and elucidated the interaction of altitude, shade and harvest periods on physical quality, preliminary cup quality, preliminary total quality, total specialty cup quality and its specific cup quality attributes (e.g. acidity, body, aroma), 100-bean weight and biochemical compositions (caffeine and chlorogenic acid content) of Ethiopian coffee beans. Caffeine and chlorogenic acid are thought to have a direct impact on human health. In addition, these two chemicals are key biochemical components of coffee beans determining its beverage quality. This study further discussed effects of altitude, shade and harvest periods on both components. Total specialty cup quality and its specific quality attributes (e.g. overall cup preference, acidity and flavour) increased both with altitude and shade level. At high altitude, however, dense shade decreased green bean cup quality and the percentage of specialty 1 (Q1) beans by ca. 50% (Table 2.8).

Coffee beans grown in cooler environments accumulate sufficient sugars and lipids (Vaast et al., 2006; Wintgens, 2004). Both are key compounds determining coffee as a beverage and are positively correlated with bean quality (Lara-Estrada & Vaast, 2007; Vaast et al., 2005). Dense shade, however, further decreased air temperature around the coffee fruits. This effect in combination with the effect of altitude on temperature could

Chapter 2

reduce growing temperature to a level below optimum, i.e. 18°C (Cortez, 1997), resulting in reduced coffee bean quality. In areas where air temperature is relatively cooler, such as at high altitudes, shade levels higher than 40 to 50% are hence not beneficial as growing temperature may decrease below optimum.

Other studies e.g.Soto-Pinto et al. (2000), indicated that shade may cause yield losses, however, high yield obtained from open sun conditions may not be consistence because high fruit load in preceding year limits bud and flower formation for the next year resulting in source limitations. This problem can be overcome via shading as it reduce fruit load through their effects on coffee physiology, such as longer internodes, fewer fruiting nodes and lower flower induction. Hence, shade grown coffee tree will give more sustainable yield than open sun grown coffee tree (Vaast et al., 2006). The presence of shade trees, especially leguminous species, improves soil fertility (organic matter content) and nutrient cycling (Kimemia, 2007; Soto-Pinto et al., 2000). Furthermore, coffee shades systems help improve coffee farmers’ revenues in the medium to long-term through diversification (timber production), marketing of high quality coffee and certification opportunities (e.g. Rainforest Alliance) (Bosselmann et al., 2009; Vaast et al., 2005).

Table 2.8: Percentage of coffee samples that are ranked as specialty 1 (Q1) and specialty 2 (Q2) grown at different altitude (high and mid), shade levels (open, medium and dense)* and harvest period (early, middle and late) ** Percentage of specialty coffee Altitude Specialty 1(Q1) Specialty 2 (Q2) High Shade level Open 53.3 46.7 Medium 66.7 33.3 Dense 33.3 66.7 Mid Shade level Open 6.7 93.3 Medium 13.3 86.7 Dense 6.7 93.3 High Harvest period Early 73.0 27.0 Middle 53.0 47.0 Late 27.0 73.0 Mid Harvest period Early 7.0 93.0 Middle 20.0 80.0 Late 7.0 93.0 *Each value is averaged across different harvest period (early, middle and late); ** each value is averaged across different shade levels (open, medium and dense).

Influence of growing altitude, shade and harvest period

The ideal growing temperature of Arabica coffee ranges between 18 and 21°C (Camargo et al., 1992) and values above or below this optimum level predisposes coffee beans to incomplete maturation and poor quality. Bosselmann et al. (2009) also reported that at high altitudes, lower temperatures due to excess shade trees resulted in reduced coffee bean quality. At mid to low altitude, in contrast, higher temperature induces the accumulation of chemical compounds such as butan-1,3-diol and butan-2,3-diol in coffee beans that are responsible for reducing quality attributes (aroma, acidity) (Bertrand et al., 2012). At high altitude, differences between open and medium shade levels on coffee bean quality were not significant. This result was in agreement with previous findings e.g. (Guyot et al., 1996) and (Bosselmann et al., 2009). This indicates that the temperature around coffee trees grown under both shade levels was still within the optimum range for coffee growth (Bosselmann et al., 2009) or allows better exposure of the trees to sunlight. At mid altitude, on the other hand, shade was observed not to have any significant effect on coffee bean quality. This might be related to leaf falling from the shade tree during the dry season, indicating that farmers should consider the best adaptable shade tree type to grow coffee under consistent shade level.

Nowadays, air temperature is also increasing due to global climate change. Besides reducing coffee bean quality, this increase in air temperature is reducing the area suitable for Arabica coffee production (Bunn et al., 2015; Davis et al., 2012). Growing coffee trees at higher altitude or under tailored shading conditions could be implemented as a possible approach to adapt to effects of climate change and to sustain production of high quality coffee beans. However, in our study area the high altitude area is limited. Hence, shading adaptation has to be considered in view of climate change.

Beans harvested at early and mid harvest period were generally higher in preliminary cup quality, preliminary total quality, total specialty cup quality, overall cup preference and body scores compared to late harvested beans. For instance, at high altitude (Table 2.8), early harvesting led to 73% of specialty 1 (Q1) coffee beans compared to 27% for late harvest. Better bean quality due to early harvest was also reported in other studies (Bertrand et al., 2005; Oberthür et al., 2011). This is possibly due to the depletion of photo-assimilates as a result of competition among fruits, which finally leads to shortage of photo-assimilates for the late developing fruits (Vaast et al., 2006). In addition to this, production of coffee quality precursor compounds, which accumulate in coffee beans are markedly high during endosperm expansion and dry matter accumulation stages

Chapter 2

(Koshiro et al., 2006). At later stage of bean development, however, most of these quality precursors are remobilized towards lignin biosynthesis and drop in relative content (Joët et al., 2009). At mid altitude, however, the harvest period did not significantly influence coffee bean quality. This is probably due to the fact that temperature at mid altitude was higher (ca. 4°C) compared to the temperature at high altitude. High temperature accelerates bean ripening and automatically resulted in a shorter time intervals between the different harvest periods.

Coffee bean weight increased with increasing altitude. The result agrees with previous studies (Vaast et al., 2006). Coffee beans grown at higher altitudes mature slowly and are, in general, harvested one to two months later than beans from mid altitudes. Slower maturation of coffee beans allows better bean filling, e.g. more lipid accumulation (Lara- Estrada & Vaast, 2007; Muschler, 2001). It is also documented that environmental factors and agricultural managements modify the physiology of coffee fruit development and ripening (De Castro & Marraccini, 2006). Among environmental factors, temperature plays a significant role in regulating bean maturation and ripening processes (Guyot et al., 1996). The present study supports the notion that low temperature slows down the ripening process (resulting in a delayed harvesting period by ca. 50 days on average) and allows better bean filling, resulting in heavier beans (Vaast et al., 2006).

Shading also promotes bean weight. Coffee beans grown under medium or dense shade are heavier than beans grown in open sun. Beans grown under shaded conditions mature slower since shading also lowers the temperature around the coffee fruits and are, in general, harvested two to four weeks later than beans in open sun (Bosselmann et al., 2009; Vaast et al., 2006). In addition, shade has also been reported to reduce the number of floral initiations per plant and to allow more assimilate partitioning to each developing bean (Bosselmann et al., 2009). Partitioning of more assimilates to individual beans increases the coffee bean weight.

However, harvest period did not significantly influence bean weight. Bean weight is mainly related to dry weight accumulation and harvest period is about bean physiological maturity and ripeness. So bean weight accumulation and harvest periods are two different stages of bean development. Bean accumulates dry matter from fruit development to bean filling. Maximum dry weight is accumulated during bean filling after which beans do not accumulate any dry matter rather starts bean maturation, colour

Influence of growing altitude, shade and harvest period

change, chemical development, tissue softening and also some flavour development. Thus, the beans harvested after maximum dry matter was accumulated and no further change in bean weight after physiological maturation. It is rather affected by shade or temperature difference not by harvest period. This could be the possible reason why bean harvesting did not affect bean weight.

In general, coffee tree grown at high altitude and under shade conditions have high leaf to fruit ratios (better bean filling capacity) which in turn leads to lower carbohydrate competitions among developing fruits. The longer leaf life span and higher photosynthetic rate of the coffee trees enhance carbohydrate supply to developing fruits (Vaast et al., 2006; Vaast et al., 2005). On the contrary, at open sun or lower altitude the life span of the leaf is short and cause lower leaf to fruit ratios. In addition, coffee tree under open sun conditions tend to flower heavily, thus producing a high fruit load without a concomitant balance in leaf area formation. Since coffee fruits act as strong sinks dry matter allocation to them may be more than four times that allocated to branch growth or leaf growth. This phenomenon reduces the photosynthesis rate of the plant and finally declines the carbohydrate supply (source limitation) to the developing fruits (Vaast et al., 2005).

Caffeine contents of our coffee samples ranged from 14.5 to 17.9 g kg-1 (Table 2.4). The results are in agreement with previous reports (Bekele, 2005; Ky et al., 2001; Silvarolla et al., 2000). But, these values varied by interaction effects of altitude, shade and harvest periods. Early harvested beans from mid altitude grown under dense shade contained the highest amount of caffeine, while the lowest caffeine content was found at high altitude, medium shading and middle harvest. In coffee bean development, biosynthesis and accumulation of most chemicals, including caffeine, is markedly higher at early development stage and reduced at later stage of ripening (Koshiro et al., 2006). At the later stage of bean development, the compounds are also remobilized towards lignin biosynthesis and their contents are relatively decreased compared to contents at early bean development stages (Joët et al., 2009; Koshiro et al., 2006). The results of the present study corroborate these findings and indicate that early harvested beans contain more caffeine than late harvested ones.

This study also showed the influence of altitude on caffeine content. An increase in altitude by 400 m decreased the caffeine content by 10% (Table 2.7). Sridevi and

Chapter 2

Parvatam (2014), reported similar findings while Avelino et al. (2005) and Bertrand et al. (2006), on the other hand, reported the opposite. Bertrand et al. (2006) showed that an increase in altitude by 500 m increased the caffeine content by 15%. In Bertrand et al. (2006), however, plots from very low altitude (e.g. 700 – 1600 m. asl.) were also included. This might be the reason for such discrepancies.

Chlorogenic acids are the main phenolic compounds in coffee beans, ranging from 6 to 12% on dry weight basis (Aerts & Baumann, 1994). In the present study, total chlorogenic acid content (sum of four subclasses) varied from 40.5 to 45.2 g kg-1 on dry weight basis (Table 2.5). These values are generally lower than values previously reported for coffee (Avelino et al., 2007; Link et al., 2014) but much higher than values reported by (Bekele, 2005; Ky et al., 2001).

In addition, we found that the values were affected by variation in altitude and shade. Total chlorogenic acid content decreased from mid to high altitude and from dense to open shading at high altitude. At mid altitude, however, coffee beans from mid altitude and open sun accumulated more chlorogenic acid. Growing temperature has a direct influence on production and accumulation of chlorogenic acids (Wintgens, 2004). The higher temperature at mid altitude probably resulted in higher chlorogenic acid production and accumulation in coffee beans grown at mid than at high altitude. A higher chlorogenic acid content indicates that coffee beans are relatively immature (Link et al., 2014).These differences in chlorogenic acid contents and bean maturation resulted in coffee bean quality differences (more Q1 coffees at high than at mid altitude). The result was in agreement with results reported by (Avelino et al., 2007; Link et al., 2014). But, studies by Bertrand et al. (2006), on the other hand, reported the opposite. In our study, the altitude ranged from 1600 – 2010 m asl. In Bertrand et al. (2006), however, plots from very low altitude (e.g. 700 – 1600 m asl.) were also included. This might be the reason for such discrepancies.

The study generally indicates that coffee growing at high altitude, in open or medium shade, and harvested at an early or middle period enhanced the potential of producing specialty 1 (Q1) coffee (Table 2.8). These growing conditions also favour the production of coffee beans with lower caffeine and chlorogenic acid content. Coffee beans with lower caffeine, but higher chlorogenic acid contents are appreciated and preferred in many consumer countries because of their positive health effect. Chlorogenic acid acts

Influence of growing altitude, shade and harvest period

as antioxidant and has radical scavenging properties. It makes coffee an acceptable beverage. It also has anti-pathogenic and allelophatic properties (Aerts & Baumann, 1994) and is described as important for disease resistance in coffee beans (Dessalegn et al., 2007).

Coffee beans grown at mid altitude had lower bean quality compared to high altitude, and were grouped mostly as Q2 specialty coffees. This indicates that there are scopes for mid altitude farmers, which occupy most of the area, to produce more Q1 coffee beans. For example, growing the trees in medium shaded conditions and selectively harvesting the beans at middle harvesting periods could potentially increase the percentage of Q1 coffees (Table 2.8).

2.5 Conclusions

Overall, the results show that growing coffee at high altitude with open or medium shade, and an early or middle harvest period increased the potential of producing beans with superior quality. Specific coffee quality attributes also increased with altitude and shade level. At high altitude, however, dense shade decreased the share of Q1 coffee by 50%. Altitude by harvest period interactions also significantly influenced coffee bean quality. At high altitude, early harvesting led to 73% Q1 beans compared to only 27% for late harvest. Regardless of harvest period, at mid altitude, coffee samples gave >80% is Q2 coffee. In general, small changes of quality attributes, driven by altitude, shading and harvest periods, can cause substantial switches in Q1 vs. Q2 classification. This quality segmentation affects coffee bean price and farmers’ competitiveness in international markets. This study, however, did not investigate year-by-year variations and soil type effects. Future studies need to integrate these impacts.

Chapter 3

Ethiopian specialty coffee quality and biochemical composition in relation to growing altitude and postharvest processing

This chapter has been submitted as

Tolessa, K., and Boeckx, P. (2016). Ethiopian specialty coffee quality and biochemical composition in relation to growing altitude and postharvest processing. Food Chemistry Journal.

57 Chapter 3

Abstract

High quality, specialty coffee is becoming a key prerequisite for reaching top-end international markets. The inherent quality and compositional characteristics of coffee beans are, however, controlled, among others, by interacting effects of growing altitude and postharvest processing techniques. The present study has been designed to examine these interactive effects on physical quality, cup quality and biochemical composition of green Arabica coffee beans. The study was carried out in Mana district, Jimma zone, Oromia regional state, south-western Ethiopia at three altitudinal ranges: high (1950 – 2100 m asl), mid (1600 – 1650 m asl) and low (1450-1550 m asl). The study generally showed significant interactions between growing altitudes and processing methods on total specialty cup quality (TSCQ), overall cup preference, aroma, aftertaste, caffeine and sucrose. Coffee beans grown at higher altitude, irrespective of the used processing methods, had higher quality scores compared to lower altitudes. The type of processing method, on the other hand, showed a clear difference on preliminary total quality in such a way that dry processed beans had higher scores than washed beans. Coffee beans grown at high altitude (ca. 87%) fall in specialty 1 grade (TSCQ ≥ 85), while all samples from mid and low altitude fall into specialty 2 grade (80 - <85). Caffeine, total chlorogenic acid and sucrose also increased with altitude. We conclude that altitude mainly controls specialty quality, but dry processing offers an option to increase coffee bean quality. In addition, because of easy management, low production cost and sustainability, dry processing is recommended over washing methods in the study area. However, to improve the quality of dry processed coffee beans, proper coffee cherry e.g. ripe red coffee cherries need to be considered.

Key words: Arabica coffee; altitude; processing method; cup quality; caffeine; chlorogenic acid; sucrose

Growing altitude and postharvest processing

3.1 Introduction

The demand of consumers for high quality coffee (specialty coffee) of single origin continues to grow (Wood, 2015). As a result, the international market asks more and more for high quality coffee of specific origin (Barbosa et al., 2014; Läderach et al., 2011). Higher quality coffees fetch higher prices on international markets as farmers are able to differentiate their product through origin and flavour (Belete et al., 2014). This market segmentation for specialty coffee also offers opportunities for coffee producing countries to generate higher incomes (Oberthür et al., 2011).

Ethiopia is known to produce a wide diversity of Coffea arabica varieties having their own unique quality attributes such as aroma and flavour. The country also has an immense potential to produce specialty coffees with a wide range of flavour characteristics (Coste et al., 1992; ICCO, 2014). Ethiopia is the largest producer and exporter of Coffea arabica in Africa (Bart et al., 2014). This commodity is Ethiopia’s number one source of foreign currency. Improving the quality is a key prerequisite for increasing export shares on international markets (Dominic & Morten, 2011; Robert, 2015). Thus, the country should sustainably produce beans with good quality attributes for which consumers are willing to pay more (Byers et al., 2008).

However, coffee quality is composed of complex traits that depend on numerous factors (Oberthür et al., 2012), among which genetic traits (Leroy et al., 2006 ), growing environment and management (e.g. shading) (Avelino et al., 2005; Decazy et al., 2003; Muschler, 2001), soil properties (Carr, 2001; Slagle et al., 2004; Yadessa et al., 2008) and postharvest processing methods (Bytof et al., 2005; Joët et al., 2010; Selmar et al., 2002) are known to predominantly affect cup quality as well as biochemical composition of coffee beans.

Coffee trees growing at high altitude produce beans with superior quality. Altitude affects coffee quality via growing temperature (Avelino et al., 2005; Decazy et al., 2003; Muschler, 2001). Lower temperature slows down bean ripening processes and allows more time for bean filling (Muschler, 2001). The ideal growing temperature of Arabica coffee is between 18 and 21°C (Camargo et al., 1992) and the temperature outside this range will cause incomplete maturation. Effect of altitude on biochemical composition of coffee beans has also been demonstrated (Avelino et al., 2007). These authors reported

Chapter 3

a higher caffeine and fat content with increasing altitude, while chlorogenic acid and sucrose contents decline. Bertrand et al. (2006), on the other hand, reported a higher chlorogenic acid content at high than at low altiude. Shade has also been found favouring coffee quality; especially under suboptimal temperature conditions (Muschler, 2001; Vaast et al., 2005a). At high altitude, however, dense shade can negatively affect coffee quality and coffee trees grown under open sun and medium shade produce beans with better quality (Chapter 2).

The quality of coffee beans not only depends on growing conditions, but also on postharvest processing methods (Joët et al., 2010; Knopp et al., 2006; Selmar et al., 2006). Dry and washed processing methods are commonly used to produce green coffee beans. Both methods remove the outer layers of coffee cherries (pericarp) and reduce the moisture content in the coffee beans to about 10–12% (Knopp et al., 2006). By dry processing, coffee cherries are directly sundried without pre-treatment and dried cherries are de-husked. By washed processing, red ripe coffee cherries are mechanically de- pulped and fermentation and washing processes remove the mucilage covering the coffee beans, before de-husking.

The semi-washed coffee processing method is another method recently introduced to Ethiopia. In this method, the fermentation step is omitted and the mucilage is removed by a de-mucilage machine. The method is known for its simplicity and improved water use efficiency compared to the fully washed coffee processing method. The washed methods without recycling uses up to 20 litres of per kilogram of cherries (Mburu et al., 1994; von Enden et al., 2002), while the semi-washed method uses about 4 litres of water per kilogram of coffee cherries and leads to a more than 80% reduction of water use compared to washing method(Bello-Mendoza & Castillo-Rivera, 1998; Matos & Monaco, 2001).

It has also been reported that dry or washed processing methods play a significant role in determining flavour characteristics of coffee beans and subsequently cause differences in the final cup quality of the beans (Bytof et al., 2005; Wintgens, 2004). It is also shown that dry processed coffee beans generate more body, whereas washed coffee beans gave better acidity and are widely accepted for their pleasant cup quality characteristics (Selmar et al., 2002). The underlying quality differences among beans processed via dry or washed methods are attributed to several metabolic activities that

Growing altitude and postharvest processing

occur in coffee beans during the course of processing (Bytof et al., 2007; Bytof et al., 2005; Kramer et al., 2010; Selmar et al., 2006).

Biochemical composition of green coffee beans, particularly the water-soluble components such as chlorogenic acid, caffeine and sugars are largely influenced by processing methods (Duarte et al., 2010; Smith, 1985). Furthermore, various biochemical compounds react and interact during coffee roasting and result in a complex mixture of flavour and aroma compounds. Hence, coffee cup quality is also linked to the biochemical composition of the green coffee beans (De Castro & Marraccini, 2006; Joët et al., 2014; Ribeiro et al., 2011).

Chlorogenic acids (CGA) are the main phenolic compounds found in green coffee beans, ranging from 6-12% (dry matter basis) (Trugo & Macrae, 1984; Van der Stegen & Van Duijn, 1980). Caffeoylquinic acids (CQA), dicaffeoylquinic acids (diCQA) and feruloylquinic acids (FQA) are three main subgroups of CGA and each individual CGA exists at least in three isomers (Clifford, 2000). CGA’s are responsible for the formation of flavour and aroma compounds in brewed coffee (De Maria et al., 1995; Farah & Donangelo, 2006) and also contribute to the final beverage acidity (Trugo & Macrae, 1984). CGA’s are known for their powerful antioxidant properties, e.g. enhanced hepatic glucose utilization (Arion et al., 1997; Herling et al., 1998), inhibition of carcinogenic compounds (Belkaid et al., 2006; Lambert et al., 2005; Stich et al., 1982) and inhibition of HIV-1 replication in cells (Robinson et al., 1996).

Caffeine is a xanthine derivative responsible for the bitter characteristic of coffee beans (Duarte et al., 2010). Recent studies also showed that moderate caffeine intake has a protective effect against cardiovascular disease and neurological disorders (Alzheimer and Parkinson disease) (O'Keefe et al., 2013; Tunson, 2014; You et al., 2011; Yukawa et al., 2004). Another study indicated that the ability of caffeine as antioxidant is the result of its degradation into methylxanthine and methylurgic acid metabolites (Lee, 2000). Owing to the antioxidant and radical scavenging properties of chlorogenic acid and caffeine metabolites, coffee is accepted as a healthy beverage (Devasagayam et al., 1996; Parras et al., 2007; Vignoli et al., 2011; Wang et al., 2015). In view of this beneficial effect on human health, these two compounds have been key subject of research (Duarte et al., 2010).

Chapter 3

Sugar, specifically, sucrose is the major component of simple carbohydrates in coffee beans. It accounts for up to 5-9% (dry matter basis) in Coffea arabica and 4-7% (dry matter basis) in Coffea canephora (Ky et al., 2001). Sucrose acts as aroma and flavour precursor and develops numerous volatile substances during roasting (e.g. furans, aldehydes and carboxylic acids), producing coffee with higher cup quality (Farah et al., 2006).

Interactive effects of growing altitude and postharvest processing methods on biochemical composition and cup quality attributes of green coffee beans are still scarce. In addition, various studies compared bean qualities of dry and washed processed coffee beans, there is limited information on the effect of semi-washed processing method on coffee bean quality. Therefore, the present study was designed to analyse physical quality, cup quality and biochemical composition of coffee beans in relation to different growing altitudes, postharvest processing methods and their possible interactions.

3.2 Materials and Methods

3.2.1 Study site

The experiment was carried out in Mana district, Jimma zone, Oromia regional state, south-western Ethiopia, during the coffee harvesting season from October 2012 to January 2013. Mana is one of the key Coffea arabica L. growing areas of Jimma zone. Its altitude ranges between 1470 and 2610 m asl, at 8°67' N and 37°07' E. The district has a total area of 47,898 ha of which 23% is at low (<1550 m asl), 65% at mid (1550- 1750 m asl) and the remaining 12% at high altitude (>1750 m asl) (JARC, 1996). The annual average temperature and rainfall are 20.5°C and 1523 mm, respectively. The soil type of the area is Nitosol (ARDO, 2008).

3.2.2 Treatments and experimental design

The experiment was carried out at three altitudinal ranges: high (1950 – 2100 m asl), mid (1600 – 1650 m asl) and low (1450-1550 m asl), with an average temperature of thirty years) of 19.3°C±1.9, 22.7°C±2.7 and 24.7°C±2.8, respectively. Fifteen coffee farms (five per altitude) with comparable coffee plant ages (7 to 10 years) were selected.

Growing altitude and postharvest processing

Agronomic management practices did not differ among the selected farms. Shade levels above the canopies of the coffee trees on each farm were quantified using SunScan canopy analysis system type SS1, BF5-RPDA1, Delta-T Devices (Cambridge, UK) and coffee trees under medium shade (40-55%) were used for cherry collection in order to avoid shade effects. Fully ripe, red coloured coffee cherries were randomly picked until 21 kg was collected per altitude and subjected to three controlled processing methods (each with 7 kg coffee cherries): dry, semi-washed and washed processing methods (see the details below). The experiment was arranged in a split plot design with three levels of altitude as a main plot and three levels of processing methods as a sub-plot with five replications (farms) resulting in a total of 45 coffee samples.

3.2.3 Processing methods . Dry processing method

Collected coffee cherries were sundried immediately on raised beds without pre- treatment. During drying the moisture content via moisture tester (mini GAC, Dickey - John, USA) of the beans was also monitored and beans were dried until a final moisture level of 11.5% was reached. The dried coffee cherries were then de-husked using a coffee hulling machine (Coffee huller, McKinnon, Scotland) and stored at room temperature prior to analysis.

Washed processing method

Red coffee cherries were mechanically de-pulped using a coffee pulper (Aagaard pregrader, McKinnon, Brazil), followed by 48 hours fermentation and mucilage removal. After washing, the coffee beans were sundried on raised beds. During drying the moisture content via moisture tester (mini GAC, Dickey - John, USA) of the beans was also monitored and beans were dried until a final moisture level of 11.5% was reached. The dried beans with parchment were de-husked using a coffee hulling machine (Coffee huller Pinhalense, Brazil) and stored at room temperature in the laboratory.

Chapter 3

Semi-washed processing method

The harvested coffee cherries were de-pulped using a mucilage remover (Pinhalense, Brazil), washed and sundried on raised beds without any fermentation process to moisture content of ca. 11.5%. The dried beans with parchment were de-husked using a coffee huller (Pinhalense, Brazil) and stored at room temperature in the laboratory.

3.2.4 Coffee quality analysis

Quality analysis was determined in laboratory of Ethiopia Commodity Exchange (ECX). The analysis was based on physical and cup quality attributes. Attributes evaluated for the physical quality were defects, odour, colour, shade and make. The attributes evaluated for cup quality were acidity, body flavour and cup cleanness. The sum of physical and cup quality scores gave preliminary total quality. All the coffee samples were qualified for specialty coffee and further analysis was done based on specific specialty cup quality attributes (see details in chapter 2, section 2.2.3).

3.2.5 Biochemical composition analysis

The biochemical composition (caffeine, chlorogenic acid and sucrose) of green coffee beans were determined. Analyses of caffeine and chlorogenic acid were performed as described by Alonso-Salces et al. (2009). Four subclasses of chlorogenic acids were identified in the green coffee beans: caffeoylquinic acid (CQA), feruloylquinic acid (FQA),dicaffeoylquinic acid (diCQA) and feruloyl-caffeoylquinic acid (FCQA).The identification was done using an already established method, liquid chromatography - time of flight - mass spectrometry (LC-TOF-MS) method (Alonso-Salces et al., 2009). Calibration was done by injecting concentration series of caffeine and chlorogenic acid every 30 samples, using an adapted response factor for the other chlorogenic acid types. Total chlorogenic acid concentration is reported as the sum of all above individual chlorogenic acids. Data acquisition and processing were done using ChromQuest 4.1 SP2 software. Detail procedures for caffeine and chlorogenic acid was presented in chapter 2 section 2.2.4.

The sucrose content of the green coffee beans was determined following Knapp (1979). First, an internal standard solution was prepared using 600 mg fenyl-beta-D-

Growing altitude and postharvest processing

glucopyranoside and 100 ml of distilled water. For each treatment, 0.5 g ground coffee beans were extracted with 25 ml distilled water containing 30 mg of fenyl-beta D- glucopyranoside as internal standard, in a hot water bath at 60°C for 30 min. Carrez I, -1 potassium hexacyanoferrate II K4Fe (CN)6 in water,(15 g L ), and Carrez II, zinc sulfate -1 ZnSO4 in water (30 g L ), were added to the extract. Distilled water was added to the extract to obtain a total volume of 50 mL. After mixing, the extract was filtered over a 0.45μm PTFE filter and 0.5 ml of the filtrate was dried using nitrogen gas in a HPLC-vial. To the dried extract 0.5 mL of “STOX” reagent (hydroxylamine hydrochloride 25 g L-1 in dry pyridine) was added and the vials were placed in an oven at 60°C for 30 min. After cooling the extract to room temperature, 0.5 mL of hexamethyldisilazane (HMDS) and 0.05 mL trifluoroacetic acid (TFA) was added and kept for 60 min to allow phase separation. Finally, 1μL was taken from the upper layer and injected into a gas chromatograph (GC-3380, Varian, USA) using an automatic injector. The detection was done using a flame ionization detector (FID).

3.2.6 Moisture content determination

The detail moisture content determination of the coffee samples is given in chapter 2, section 2.2.5

3.2.7 Data analysis

Data were analysed with Statistical Analysis System software (v. 9.2, SAS Institute Inc., Cary, NC USA) using a mixed model procedure for a split plot design with altitude as the main ‘plot’ and processing methods as ‘sub-plot’. Significant differences between treatment means were determined using Tukey’s, honest significant difference, HSD test.

3.3 Results

3.3.1 Coffee quality

Significant interactions between altitudes and processing methods were only found for total specialty cup quality and its specific specialty coffee quality attributes: overall cup preferences, aroma and aftertaste (P < 0.01; Table 3.1). For all processing methods, the

Chapter 3

scores for each of these attributes increased with altitude. At the highest altitude, scores for overall cup preference (8.4) and aroma (8.0) were, however, higher for dry processed beans than for semi-washed or washed beans (Table 3.1).

Altitude as a main effect significantly increased scores for most quality attributes. Preliminary cup quality (50.6), preliminary total quality (85.4), total specialty cup quality (85.2), overall cup preference (8.1), acidity (7.9), body (7.8), aroma (7.9), flavour (8.1) and aftertaste (7.8) had a higher score at high than at mid and low altitude. The scores for these attributes at mid and low growing altitude, on the other hand, were not significantly different from each other (Table 3.2). Growing altitudes did not generally influence the physical quality of coffee beans (P > 0.01; Table 3.2). Processing methods, on the other hand, significantly affected the physical quality and preliminary total quality (P < 0.01; Table 3.2). The quality scores of these attributes were higher for dry processed coffee beans than for washed processed beans (Table 3.2). Though, the differences were not robust, similar trends were also observed for overall cup preference, aroma and aftertaste.

Growing altitude and postharvest processing

Table 3.1: Interactive effects of altitude (A) and processing methods (PM) on the physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and the specific specialty cup quality attributes score; Altitude: High (1950 – 2100 m asl), Mid (1600 – 1650 m asl) and Low (1450-1550 m asl); processing methods: Dry (DP), Semi-washed (SW) and Washed (W). A PM PQ PCQ PTQ TSCQ Specific specialty cup quality attributes OCP1 Acidity Body Aroma Flavour AT2 a a a a High DP 38.2 ± 0.8 50.4 ±1.0 88.6 ±1.0a 86.6 ± 0.9a 8.4 ± 0.1a 8.0 ±<0.1a 7.9 ± 0.1a 8.0 ± 0<0.1a 8.1± 0.1 7.8 ±<0.1 SW 33.4 ± 0.9a 51.0 ±1.2a 84.4 ± 0.9a 85.8 ± 0.3a 7.9 ± 0.1b 7.9 ±<0.1a 7.8 ±<0.1a 7.9 ± 0.1b 8.1± 0.1a 7.9 ± 0.1a WP 32.8 ± 0.3a 50.4 ±1.0a 83.2 ± 0.9a 85.0 ± 0.3a 7.9 ± 0.1b 7.9 ±<0.1a 7.8 ±<0.1a 7.8 ±<0.1b 8.1± 0.1a 7.7 ±<0.1a Mid DP 37.0 ± 0.9a 46.2 ± 0.7a 83.2 ± 0.7a 82.5 ± 0.3c 7.5 ± 0.1c 7.7 ±<0.1a 7.6 ± 0.1a 7.5 ±<0.1cd 7.5 ±<0.14a 7.3 ±<0.1c SW 35.0 ± 0.8a 48.6 ±1.0a 83.6 ± 0.5a 83.6 ± 0.3b 7.8 ± <0.1b 7.8 ±<0.1a 7.6 ± 0.1a 7.6 ±<0.1c 7.7 ±<0.1a 7.7 ± 0.1a WP 34.8 ± 1.0a 46.2 ± 0.7a 83.0 ± 0.8a 83.2 ± 0.2b 7.7 ± 0.1b 7.8 ± 0.1a 7.5 ± <0.1a 7.5 ±<0.1cd 7.7± 0.1a 7.5 ±<0.1b Low DP 37.0 ± 1.1a 46.2 ± 1.4a 83.2 ±1.4a 83.1 ± 0.2b 7.7 ± 0.1b 7.8 ± 0.1a 7.6 ± 0.1a 7.5 ± 0.1cd 7.7 ± 0.1a 7.6 ±<0.1b SW 33.0 ± 1.3a 48.0 ± 0.9a 81.0 ±1.2a 83.0 ± 0.3b 7.5 ± 0.1c 7.7 ± 0.1a 7.6 ± 0.1a 7.5 ±<0.1cd 7.5 ±<0.1a 7.6 ±<0.1b WP 32.8 ± 0.7a 46.8 ± 0.7a 79.6 ±1.3a 82.6 ± 0.3c 7.4 ± 0.1d 7.7 ± 0.1a 7.6 ± 0.1a 7.5 ±<0.1cd 7.6 ± 0.1a 7.5 ±<0.1b P-value 0.23 0.90 0.34 0.0075 < 0.0001 0.38 0.95 0.0013 0.057 0.008 1Overall cup preference; 2Aftertaste; different letters in the same column indicate significant difference according to Tukey’s HSD post hoc test (P < 0.01); results are shown as mean ± standard error

Table 3.2: Effects of altitude (A) and processing methods (PM) on the physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty cup quality attributes scores; Altitude: High (1950 – 2100 m asl), Mid (1600 – 1650 m asl) and Low (1450 – 1550 m asl); processing methods: Dry (DP), Semi – washed (SW) and Washed (W) Factor level PQ PCQ PTQ TSCQ Specific specialty cup quality attributes OCP1 Acidity Body Aroma Flavour AT2 A High 34.8 ± 0.7a 50.6 ± 0.6a 85.4 ± 0.8a 85.2 ± 0.7a 8.1 ± 0.1a 7.9 ±<0.1a 7.8 ±<0.1a 7.9 ±<0.1a 8.1±<0.1a 7.8 ±<0.1a Mid 35.6 ± 0.5a 47.0 ± 0.5b 82.6 ± 0.5b 83.1 ± 0.2b 7.6 ± 0.1b 7.7 ±<0.1b 7.6±<0.1b 7.5 ±<0.1b 7.6 ±<0.1b 7.8 ± 0.1b Low 34.3 ± 0.7a 47.0 ± 0.6b 81.3 ± 0.8b 82.9 ± 0.2b 7.5 ±<0.1b 7.7 ±<0.1b 7.6± 0.1b 7.5 ±<0.1b 7.6 ±<0.1b 7.5 ±<0.1b P-value 0.30 0.0018 0.0018 0.015 <0.0001 <0.001 <0.001 <0.001 <0.0001 <0.0001 PM DP 37.4 ± 0.3a 47.6 ± 0.8b 85.0 ± 0.9a 83.2 ± 0.7a 7.8 ± 0.10a 7.8 ±<0.1a 7.7 ± 0.1a 7.7 ± 0.1a 7.7 ± 0.1a 7.6 ± 0.1b SW 33.8 ± 0.6b 49.2 ± 0.7a 83.0 ± 0.6b 84.1 ± 0.3a 7.8 ± 0.1a 7.8 ±<0.1a 7.7 ± 0.1a 7.7 ± 0.1a 7.7 ± 0.1a 7.7 ±<0.1a W 33.5 ± 0.5b 47.8 ± 0.7b 81.3 ± 0.7b 83.5 ± 0.3a 7.7 ± 0.1b 7.8 ±<0.1a 7.6 ± 0.1a 7.6 ±<0.1b 7.8 ± 0.1a 7.5 ±<0.1b P-value <.0001 0.016 0.0014 0.34 0.003 0.58 0.60 0.011 0.44 <0.001 1Overall cup preference; 2Aftertaste, different letters in the same column indicate significant difference according to Tukey’s HSD post hoc test (P < 0.01); results are shown as mean ± standard error

Chapter 3

3.3.2 Caffeine, total chlorogenic acid and sucrose content

Significant interactions between altitudes and processing methods were also observed for caffeine and sucrose content (Table 3.3, P < 0.01). Total chlorogenic acid, on the other hand, was not influenced by this interaction (P > 0.01). Coffee beans produced at higher altitude and processed via the semi-washed processing method had the highest sucrose content (90.5 g kg-1) compared to beans from mid and low altitude processed by any of the processing methods (Table 3.3). Regardless of processing method, coffee beans from low altitude, compared with beans from high and mid altitudes, generally had the lowest caffeine content.

Table 3.3: Interactive effects of altitude (A) and processing methods (PM) on caffeine, total chlorogenic acid and sucrose contents (g kg-1 dry weight) of green coffee beans; Altitude: High (1950 – 2100 m asl), Mid (1600 – 1650 m asl) and Low (1450 – 1550 m asl); processing methods: Dry (DP), Semi – washed (SW) and Washed (W) A PM Caffeine Total chlorogenic acid Sucrose High DP 13.5 ± 0.2ab 62.9 ± 0.9a 86.5 ± 0.6b SW 13.8 ± 0.2a 67.4 ± 2.6a 90.5 ± 1.1a W 13.6 ± 0.2ab 63.3 ± 1.5a 82.3 ± 1.0cde Mid DP 12.9 ± 0.9bc 59.8 ± 1.1a 82.4 ± 0.8cde SW 13.8 ± 0.8a 62.5 ± 1.1a 84.7 ± 0.4cd W 13.8 ± 0.9a 59.3 ± 1.2a 83.9 ± 0.4bc Low DP 12.2 ± 0.3c 57.3 ± 0.9a 79.1 ± 1.1e SW 12.0 ± 0.2c 61.0 ± 0.2a 80.8 ± 1.1cde W 11.9 ± 0.2c 57.9 ± 1.7a 79.3 ± 1.4ed P-value 0.009 0.96 <0.0001 Different letters in the same column indicate significant difference according to Tukey’s, HSD post hoc test (P < 0.01); results are shown as mean ± standard error

Table 3.4: Effects of altitude (A) and processing methods (PM) on caffeine, total chlorogenic acid and sucrose contents (g kg-1 dry weight ) of green coffee beans; Altitude: High (1950 – 2100 m asl), Mid (1600 – 1650 m asl) and Low (1450 – 1550m asl); processing methods: Dry (DP), Semi – washed (SW) and Washed (WP) Factor Level Caffeine Total chlorogenic acid Sucrose A High 13.6 ± 0.1a 64.5 ± 1.2a 86.4 ± 1.0a Mid 13.5 ± 0.1a 60.5 ± 0.8b 83.7 ± 0.4b Low 12.0 ± 0.1b 58.7 ± 0.9b 79.7 ± 0.7c P-value <0.0001 0.009 0.007 PM DP 12.8 ± 0.2b 60.0 ± 0.8b 82.7 ± 0.9b SW 13.2 ± 0.2a 63.7 ± 1.1a 85.1 ± 0.9a W 13.1 ± 0.2ab 60.1 ± 1.1b 82.1 ± 0.8b P-value 0.024 0.006 <0.0001 Different letters in the same column indicate significant difference according to Tukey’s HSD post hoc test (P < 0.01); results are shown as mean ± standard error.

Growing altitude and postharvest processing

Main effects are given in Table 3.4. Altitude significantly influenced caffeine, total chlorogenic acid and sucrose content. Coffee beans produced at high altitude had higher caffeine (13.6 g kg-1), total chlorogenic acid (64.5 g kg-1) and sucrose (86.4 g kg-1) content compared to the lower altitudes. Processing method as main effect also affected the total chlorogenic acid and sucrose content but not the caffeine content. The highest total chlorogenic acid (63.7 g kg-1) and (85.1 g kg-1) sucrose content was observed for the semi-washed processing method.

3.4 Discussion

High quality is a major attribute for the coffee industry as it is the basis for success in a strong competitive market (Barbin et al., 2014). Production of high quality coffee is mainly associated with growing environment, soil properties, genetic traits and postharvest processing methods. A number of studies evaluated influences of altitude (Avelino et al., 2005; Avelino et al., 2007; Bosselmann et al., 2009; Decazy et al., 2003) and postharvest processing methods (Joët et al., 2010; Knopp et al., 2006; Selmar et al., 2002) on cup quality and biochemical composition of green coffee beans. Information on the interactive effects of both factors on quality and composition of green beans is, however, not widely available. Therefore, we studied interaction effects of altitude and postharvest processing methods on physical quality, cup quality attributes and selected biochemical composition of Ethiopian coffee beans. We focused on sucrose, caffeine and chlorogenic acid as main biochemical composition because these attributes are the dominant precursors for coffee aroma and flavour development. The latter two also have a positive impact on human health.

For the three processing methods, scores of most coffee quality attributes increased with increasing altitude. These results are in agreement with previous studies (Avelino et al., 2005; Avelino et al., 2007; Bosselmann et al., 2009; Decazy et al., 2003). The positive effects of altitude on coffee bean quality are mainly due to lower temperature at higher compared to lower altitude. Low temperature promotes production of beans with better quality (Vaast et al., 2006). Lower temperature (in our study area a difference of ca. 5°C between high and low altitude), in general, slows down the ripening process and enhances accumulation of aroma precursors, such as fat and carbohydrates (Vaast et al., 2006). These chemical components of coffee beans have been reported to have a

Chapter 3

positive and strong correlation with coffee bean quality and its cup quality attributes (Lara-Estrada & Vaast, 2007; Ribeiro et al., 2011).

Recently, it has also been shown that coffee production is sensitive to climate change (Bunn et al., 2015). Due to an increasing temperature, coffee producers at low altitude may face problems to sustain green bean quality. Thus, modification of growing environment, e.g. using shading trees, will help these producers to guarantee production of high quality coffee beans. However, a shade percentage of more than 50%, on the other hand, may not be beneficial and probably reduce coffee quality. This affects farmers’ decision and forces coffee production to shift from low to high altitude. However, shortage of suitable area and land topography at high altitude, on the other hand, may stop further expansion of coffee production.

Regardless of altitude, the dry processing method improved the quality score of preliminary total quality (PTQ), which may increase the percentage of coffee samples to be graded as specialty coffee (PTQ > 80). In addition, dry processing also significantly enhanced the quality scores of overall cup preference and aroma at high compared to low and mid altitude. These results are in agreement with previous reports from e.g. Bytof et al. (2005) and Selmar et al. (2002). These authors reported significant differences in coffee bean quality and chemical compositions between dry and washed processed coffee beans. The results could be attributed to the difference in metabolic processes that occur during processing (Bytof et al., 2007; Bytof et al., 2005; Kramer et al., 2010; Selmar et al., 2006).

Compared to semi-washed and washed processing methods, the dry processing method is the simplest, cheapest and has the largest processing capacity (10,000 kg h-1). Dry processing is also known for its eco-friendly nature, while the other methods are criticized not to preserve other ecosystem services due to their large water usages, ca. 20 litre of water per kilogram of coffee cherry. From an environmental, economic and coffee quality perspective point of view and inexpensiveness, this study further promotes dry processing method for coffee growers in the study area as it 1) increase the change to reach specialty quality threshold scores, 2) has a 10-fold lower processing cost, and 3) and preserves more other ecosystem services. However, to achieve the highest quality coffee via dry processing, appropriate coffee cherry e.g. ripe red coffee cherries need to be considered. In addition due attention should also be given to appropriate drying

Growing altitude and postharvest processing

material e.g. raised bed with mesh wires to enhance the coffee quality of dry processing method (Tsegaye et al., 2014).

Caffeine, total chlorogenic acid and sucrose contents of our coffee samples ranged from 11.9 to 13.8 g kg-1, 57.9 to 67.4 g kg-1 and 79.3 to 90.5 g kg-1, respectively and the values increased with increasing altitude (Table 3.3). However, these results are in contrary to our previous result (Chapter 2), where the contents of caffeine and total chlorogenic acid decreased with altitude. This negative influence of altitude on caffeine content is also contradicted with the other previous studies (Avelino et al., 2005; Avelino et al., 2007; Link et al., 2014) but agrees with Sridevi and Parvatam (2014). Bertrand et al. (2006), on the other hand, reported both negative and positive effect of altitude on caffeine and chlorogenic acid. For instance, the contents of these components increased with altitude ranges between 900-1400m but decreased at 1450m. Borém et al. (2014) also reported higher level of chlorogenic acid for the coffee grown above 1200 m.

Still there is no consensus in literature about the effect of altitude on these components, however, the reason for such discrepancies might be due the variability of factors considered during the studies rather than altitude. For instance, in our previous study (Chapter 2), apart from altitude, different shade levels (open sun, medium shade and dense shade) and different harvest periods (early, middle and late) were considered while in chapter 3 only medium shade condition and early harvest along with altitude were considered. Moreover, production year or season might be also other reason since coffee plant does not give the same yield and quality which can be explained by variation of climatic conditions i.e. temperature and rain fall and the biennial bearing nature of the coffee plant. Biennial bearing (alternate bearing) means that years with high fruit yields alternate with years with low fruit yield (DaMatta, 2004).

The chemical composition of coffee beans varies with their degree of maturity and ripening. Coffee maturity and ripening, on the other hand, are influenced by the growing environment (Guyot et al., 1996). Chlorogenic acid and related chemical contents were reported to increase from green to red-violet matured beans (Clifford & Kazi, 1987). The metabolism of bean compounds (e.g. chlorogenic acid) was also found to be controlled by growing temperature (Joët et al., 2010). Higher temperature accelerates the maturity of coffee beans and leads to incomplete bean filling. In this condition many metabolic pathways of the chemical composition such as chlorogenic acid are incomplete. This is

Chapter 3

possibly due to the high temperature negatively affecting enzyme activities that involve in biosynthesis of chlorogenic acids which in turns reduce their final accumulation.

Coffee beans grown at higher altitudes, on the other hand, mature slowly and are harvested ca. two months later than beans from low altitude. This allows better bean filling and better accumulation of aroma precursors. It has also been noted that coffee trees grown at higher altitude have a higher leaf to fruit ratio than at lower altitude (Vaast et al., 2005). This leads to an increased supply of photo-assimilates to the developing beans and resulting in a higher accumulation of chemical composition.

In the present study, coffee trees grown at mid and at low altitude had more fruits per tree than trees at high altitude (data not shown). This reinforces the notion that temperature strongly influences rates of floral and fruit development (DaMatta, 2004). According to Bote and Vos (2016), more fruits per tree enhance competition for assimilates among coffee beans. A strong competition disproportionately limits accumulation of chemical components in coffee beans.

The caffeine content was not significantly influenced by different processing methods and the result agrees with the existing information in the literature (Balyaya & Clifford, 1995; Duarte et al., 2010; Leloup et al., 2005). However, semi-washed coffee beans had a higher total chlorogenic acid and sucrose content than dry or washed processed beans. Since chlorogenic acid and sucrose are relatively soluble in water (Smith, 1985; Van der Stegen & Van Duijn, 1980), a higher amount of these compounds is leached out of the beans during fermentation of washed processed beans. However, differences in biochemical composition among dry, semi-washed and washed processed coffee beans were not substantial to allow a shift from specialty coffee 2 (Q2) into Q1 beans. This indicates that growing altitude is the main driving factor to control coffee bean quality development.

This study generally reveals that growing coffee at high altitude improved coffee bean quality and biochemical composition. At this growing altitude, about 87% of coffee samples were graded in specialty 1 (Q1), while coffee grown at mid and low altitude fall into Q2 grades (data not shown). Caffeine, total chlorogenic acid and sucrose content also increased with altitude by ca. 11, 9 and 10%, respectively (Table 3.4).

Growing altitude and postharvest processing

3.5 Conclusions

In general, this study confirmed that growing altitude is the main factor to control coffee bean quality. Almost all coffee beans from high altitude fall into Q1 grade. The biochemical composition e.g. chlorogenic acid and sucrose of green coffee beans also increased with altitude. Thus, higher growing altitudes, in general, have potential to produce high quality and healthier coffee beans. Though differences in quality among dry, semi-washed and washed processed coffee beans were not consistent, dry processed coffee beans gave the highest scores for preliminary total quality, overall cup preference and aroma. Moreover, dry processing is cheaper and preserves other ecosystem services more than washed processing methods. However, to enhance the quality of dry processed coffee beans, appropriate coffee cherry e.g. ripe red coffee cherries need to be considered.

Chapter 4

Analysis of coffee quality along the coffee value chain in Jimma zone, Ethiopia

This chapter is to be submitted as

Tolessa, K., and Boeckx, P. (2016). Analysis of coffee quality along the coffee value chain in Jimma zone, Ethiopia to Journal of food quality.

Chapter 4

Abstract

This study describes the effect of cooperative, certification, private trader, farmers, sorting and postharvest processing methods on Arabica coffee quality. Coffee samples were collected from certified cooperatives, non-certified cooperatives, private traders and farmer’s, member of certified cooperatives, non-certified cooperatives and non-member of a cooperative, who sell to private traders; all belonging to three coffee growing districts of Jimma zone, Ethiopia. The study showed that coffee beans sampled from cooperatives had higher quality scores and were all classified as either specialty 1 (Q1) (33%) or specialty 2 (Q2) (67%). About 78% of coffee beans sampled from private traders fall in the commercial coffee grade (grade 3), while only 22% of their beans qualified for Q2. Coffee certification, in general, did not add any value to coffee quality. No quality differences were also observed between coffee beans sampled from farmers that were members of cooperatives and farmers that were non-members of the cooperatives. Coffee quality differences were observed between coffee processing methods. In such a way that dry processing method improved quality scores of physical quality, preliminary total quality total speciality cup quality, overall cup preference and acidity. However, this can only be achieved by using appropriate coffee cherries i.e. ripe red and clean coffee cherries. Moreover, based on its cost effectiveness and eco-friendly nature, our study recommends dry processing method in the study area. The study also showed sorting of coffee cherries improved the quality of coffee and increased the percentage of coffee samples qualifies for Q1 coffees. In general proper coffee cherries type along with site specific coffee processing approach help the coffee actors to produce high quality coffee. However, the study did not consider different seasons. This therefore requires further studies to substantiate this recommendation.

Key words: Arabica coffee; cooperatives; private traders; certification; famers

Analysis of coffee quality along the value chain

4.1 Introduction

Coffee is the world’s favourite beverage and most traded commodities (Barbosa et al., 2014; Davis et al., 2012; Murthy & Naidu, 2012). Because of the increasing demand for high quality and product differentiation from the consumer side (Curzi et al., 2014), coffee quality becomes an important attribute on the international coffee market. This increasing demand for high quality coffee triggered coffee producing countries to produce high quality, specialty coffee for high - end markets. Ethiopia is known for the origin and wide diversity of Arabica coffee and has enormous, unexplored potential to produce top specialty coffees (Anthony et al., 2001; Coste et al., 1992). The country is Africa's leading producer and exporter of Arabica coffee (Tefera, 2015). Coffee production is the backbone of the Ethiopian economy, contributing 25-30 per cent of total export earnings (Tefera & Tefera, 2014). Coffee further plays a major role in sustaining the livelihoods of more than 15 million coffee growing households in the country (Davis et al., 2012). Majority of coffee produced in the country (95%) is grown by smallholders farmers while the remaining 5% comes from large coffee plantations (Tefera, 2015).

Coffee produced by these smallholder farmers reaches consumers by passing through different value chain actors. The actors at each stage add value to the final product. In Ethiopia, coffee value chain actors include: input providers, producers (smallholder farmers, private growers and coffee plantation), private trader, cooperatives, unions (association of cooperatives), Ethiopia Commodity Exchange (ECX), various government institutions, exporters and finally the consumer (Gemech & Struthers, 2007). Cooperative is an alternative options of Ethiopian coffee value chain to enhance the competitiveness of the smallholder coffee farmers through modernization of coffee production and marketing (Bernard & Spielman, 2009; Dorsey & Assefa, 2005). The cooperatives work directly with farmers and provide them: processing machines, accesses to credit, transportations, training and price information. Cooperatives enable farmers to improve coffee quality and ultimately their income (Dempsey & Campbell, 2006; Kodama, 2007).

Chapter 4

Unions are associations of cooperatives and were established with the aim of enhancing the economic scale through increasing the bargaining power of cooperatives in selling their product (Emana, 2009; Meijerink et al., 2010). Cooperative unions are allowed to sell their coffee directly to international buyers (Tefera, 2016). As a result, unions took the lead to establish international market linkage and facilitate direct export on behalf of smallholder coffee producers. This linkages strengthen and encourage coffee growers to produce quality product, efficient operations and market synergy (Dempsey, 2006). Duromina coffee cooperative in Jimma zone, member of Oromia coffee union, for instance, improved the quality of their coffee beans and received premium price i.e. ca $8.20 per kilogram of green coffee beans in 2012. Hence, the “new” coffee value chain: cooperatives and unions; have gained particular importance on international specialty coffee market (Stellmacher, 2011) and increased local farmers’ market share in the international market (Emana, 2009; Kodama, 2007).

In consumer countries, in addition to quality, there is also a growing demand for healthier and eco-friendly produced coffee (Giovannucci et al., 2014; Jena et al., 2012; Stellmacher, 2007). Finally, certification has become an important tool for differentiations and creation of niche markets for such coffee products (Grote et al., 2007) (Daniele, 2008). Certification also guarantees foreign buyers about the origin and quality of the product and offers producers the potential to receive premium price for their certified coffee beans (ITC, 2012). But there are some challenges which may limit the function e.g. weak organization capacity of the cooperatives, lack of awareness about the certification and high of certification cost.

In Ethiopia, the common coffee certification types include organic (i.e. coffee grown without chemical fertilizer and pesticides), fair trade (promotes greater economic incentives via fair price and direct trade), rainforest alliance (integrates biodiversity conservation, community development and suitable agricultural practices to ensure sustainable farm management) and UTZ certification (provides market transparency and traceability through promoting good agricultural practices at the farm level (Adam, 2009; Potts et al., 2014; Volkmann, 2008). All these certifications are expected to offer a combination of benefits to smallholder coffee farmers including increased market access, higher and stable income (Ruben & Fort,

Analysis of coffee quality along the value chain

2012). In some other countries e.g. Northern Nicaragua, smallholder coffee producers and certified cooperatives significantly benefited compared to non-certified farmers and cooperatives (Bacon, 2005; Dorr et al., 2010; Philpott et al., 2007; Poncelet et al., 2005).In addition the demand for certified coffee has risen alongside the demand for specialty coffee (ITC, 2012). Despite the increasing demand for certification and establishment of a “new” coffee value chain in Ethiopia, there are so far no studies that report impacts of cooperatives and certification on coffee bean quality.

Postharvest processing method is another key factor that influence the final coffee quality and chemical compounds (Bytof et al., 2005; Duarte et al., 2010; Knopp et al., 2006; Selmar et al., 2002). In Ethiopian, about 70 - 80% is dry processed coffee beans while the remaining 20 - 30% are washed processed coffee beans (Bart et al., 2014). It is generally believed that highest quality coffees in the country are obtained via washed processing method and receive higher premium than dry processed beans. For the washed processing method only red ripe cherries are used while for dry processing method most of the time coffee cherries of different ripening stages e.g. unripe, ripe and overripe are used. This might be the main reason affecting qualities of dry processed coffee beans. Almost all cooperatives are using the washed processing method to produce high quality coffee. However, this method is expensive, consume large amounts of water and pollute the environment. Hence, this need further research to investigate the impacts of processing (dry and washed methods) on coffee quality by considering identical coffee cherries for further intervention. It is also important to assess the coffee cherry quality at farmer’s level. Because the quality of the coffee beans produced via any of processing methods depend on the type of coffee cherries at harvest e.g. unripe, red ripe and overripe.

This study therefore, was aimed first to determine coffee quality of coffee beans from certified cooperatives, non-certified cooperatives, and private traders. Secondly to assess the variability of coffee quality among different coffee farmers who are a members cooperatives (certified and non-certified cooperatives) and non- member cooperatives, who sell to private traders processed via dry and washed processing methods. The study was also examined the impact of sorting coffee cherries on the coffee bean quality of farmers

Chapter 4

who are a members cooperatives (certified and non-certified cooperatives) and non- member cooperatives, who sell to private traders.

4. 2 Materials and Methods

4.2.1 Study site

The experiment was carried out in three districts (Mana, Goma and Limu) of the Jimma zone, Oromia regional state of Ethiopia. They are known as predominant Coffea arabica growing areas. Mana district is located at an elevation of 1400 - 2610 m asl, 7° 67' N, 37° 07' E with mean annual temperature and rainfall of 20.5°C and 1525 mm, respectively. The soil of the area is characterized as a Nitisol, with pH ranging from 4.5 to 5.5 (ARDO, 2008). Goma district is located at an elevation of 1400 - 2270 m asl, 7° 57' N, 36° 42' E, with annual mean temperature and rainfall of 21.7°C and 1600 mm, respectively. The soil of the area is characterized as an Eutric Nitisol, with pH ranging from 4.5 – 6.0 (IPMS, 2007). The third site, Limu, is located at an elevation of 1600 - 1800 m asl, 8° 05' N, 36° 57' E, with an annual mean temperature and rainfall of 20.4°C and 1616 mm, respectively. The soils are dominated by Eutric Nitisols with pH ranging from 4.5 - 5.8.

4.2.2 Treatments and experimental design

For this study, coffee actors along the value chain (cooperatives, private traders and farmers) of three districts (Mana, Goma and Limu) were considered. The study was composed of three experiments. In experiment 1, washed processed coffee beans were sampled from cooperatives: certified cooperatives (CC) and non- certified cooperatives (NCC) and private traders (PT). In experiment 2, red coffee cherry samples were collected from farmers: members certified cooperatives (FMCC), member non-certified cooperatives (FMNCC) and non-member of cooperatives (FNMC) i.e. farmers who sell to private traders and processed via dry and washed processing methods without sorting. In experiment 3, red coffee cherry samples were also collected from farmers; members certified cooperatives (FMCC), member non-certified cooperatives (FMNCC) and non-member of cooperatives (FNMC) and

Analysis of coffee quality along the value chain sorted out. The sorted out coffee cherries were subjected two processing methods (dry and washed methods). In all experiments, coffee samples were collected from October 2012 to February 2013. Detailed description of each experiment is presented below.

Experiment 1

Coffee cooperative offices of each district were contacted to collect information on number of cooperatives and their organizational structure. Based on the obtained information six coffee cooperatives (three certified and three non-certified) were selected from Mana and Goma district. In these districts, only three certified cooperatives were present. In Limu district, all cooperatives were certified and three certified cooperatives were randomly selected. Additionally, nine private traders, three from each district, were randomly selected. Certified and non-certified cooperatives buy coffee cherries from their own respective members while private traders buy from farmers who were not a member of cooperatives via their agents i.e. village assemblers. Both private traders and cooperatives process the coffee cherries by themselves.

For coffee quality analysis a composite sample of one kilogram of washed processed green coffee beans was collected. This composite sample was taken from on average from 30 bags (i.e. 100 kg per bag) containing green coffee beans. The experiment was arranged in a “split plot” design, with district as a “main plot” and three levels of actors (certified cooperatives, non certified cooperatives and private traders) as a “sub-plot”.

Experiment 2

The aim of this experiment was to examine the variability of coffee quality among farmers who are a members certified cooperatives (FMCC), member non-certified cooperatives (FMNCC) and non-member of cooperatives (FNMC) and processing method; dry vs. washed. From each three districts, six farmers were randomly selected and coffee cherries were collected from local market places where farmers sell cherries to either cooperatives or private traders. The collected samples (14 kg from each farmer) were subjected to two processing methods (dry and washed) at Goma I and Limu washing stations. These coffee 81

Chapter 4

samples were processed without sorting out low or bad quality coffee cherries (e.g. unripe and overripe cherries). This allows assessing the actual coffee harvesting practice of farmers towards coffee quality. The experiment was arranged in a “split-split plot” design with districts as a “main plot”, farmer type as “sub-plot” and processing type (washed and dry processing) as a “sub-sub-plot”.

Experiment 3

In this experiment coffee cherries collected from farmers who are member and non-member of cooperatives were sorted for unripe and overripe coffee cherries before processing. The main aim was to investigate coffee quality after sorting is applied before processing. For this, only Goma district was considered. Coffee cherries (28 kg per farmer) were collected at the local market from six farmers per type. Half of the cherries (14 kg) were sorted for unripe and overripe cherries and only clean and ripe coffee cherries were subjected to dry and washed processing. The other 14 kg of unsorted coffee cherries were also subjected dry and washed processing. The experiment was arranged in a “split-split plot” design with actors (farmers, member of certified and non-cooperatives, non-member cooperatives) as a “main plot”, sorting treatment as a “sub-plot” and processing methods (washed and dry processing, see the details in chapter 3 section 3.2.3) as “sub-sub-plot”.

4.2.3 Coffee quality analysis

Coffee quality was assessed based on both the physical and cup quality which allows to give preliminary total quality score that defines the quality of the coffee samples. Here cup tasting was performed by a team of three experts who operate commercially at ECX. For the scores higher than 70; cup quality for specialty coffees were further assessed for overall cup preference, acidity, body, aroma, flavour, aftertaste, uniformity, cup cleanness, sweetness and balance (see details in chapter 2, section 2.2.3).

Analysis of coffee quality along the value chain

4.2.4 Data analysis

Data were analysed with Statistical Analysis System software (v. 9.2, SAS Institute Inc., Cary, NC USA) using a mixed model procedure for a split plot design. Significant differences between treatment means were determined using Tukey’s, honest significant difference, HSD test.

4.3 Results

4.3.1 Coffee quality

The result of experiment 1 (Table 4.1) indicated that different coffee actors significantly influenced (P < 0.01) preliminary cup quality, preliminary total quality, total specialty cup quality, acidity and aroma of coffee beans. Coffee beans of non-certified cooperatives had higher quality scores for preliminary cup quality (50.0) and preliminary total quality (77.7) than beans from certified cooperatives and private traders (Table 4.1). Total specialty cup quality of certified (82.3) and non-certified cooperatives (84.3) was higher compared to private trader’s (78.8) coffee. Acidity and aroma scores were also higher for certified (7.7 and 7.4) and non-certified (8.1 and 7.7) cooperatives compared to private traders (7.2 and 6.9), respectively (Table 1). Different coffee growing districts, on the other hand, did not show any significant effect for all coffee quality attributes. In experiment 2 (Table 4.2), interactions between actors and processing methods significantly affected overall cup preference (P < 0.01). Dry processed coffee beans sampled from non-certified cooperatives gave the higher quality scores compared to beans from any other combinations (Table 4.3).

Main effects of experiment 2 are shown in Table 4.3. As a main effect, district had a significant effect on preliminary cup quality, total specialty cup quality and coffee bean flavour characteristics. For these quality attributes, coffee beans sampled from Limu had the highest scores while coffees from Mana had the lowest. Similarly, dry processed coffee beans had higher physical quality, preliminary total quality, total specialty cup quality and specific specialty cup quality scores than washed processed coffee beans. Preliminary cup 83

Chapter 4

quality, on the other hand, was not significantly affected by processing methods (P > 0.01; Table 4.3). In addition, no significant effect of actors was observed for all bean quality attributes (P > 0.01; Table 4.3).

The result of experiment 3 (Table 4.5) showed that cherry sorting significantly affected (P < 0.01) for all coffee quality attributes except preliminary cup quality. The highest quality scores were obtained from sorted coffee beans for physical quality (38.3), preliminary total quality (85.3) total specialty cup quality (85.2), overall cup preference (7.9), acidity (7.9), body (7.9), aroma (7.8), flavour (7.8) and aftertaste (7.7) compared the unsorted coffee beans (Table 4.4).

Coffee quality attributes were also significantly influenced by processing method (P < 0.01). Dry processing methods enhanced quality scores of physical quality (38.9), preliminary total quality (85.1), total specialty cup quality (84.8) and specific quality attributes: overall cup preference (7.9) and acidity (7.9). Processing methods, on the other hand, did not significantly affect preliminary cup quality, body, aroma, flavour and aftertaste (P > 0.1). However, the differences observed among farmers were not significant (P > 0.01; Table 4.4The three and two - way interactions between farmers, sorting treatment and processing methods for all coffee quality attributes were also non-significant (data not shown).

Analysis of coffee quality along the value chain

Experiment 1: Coffee quality analysis along the value chain in Mana, Goma and Limu districts of Jimma zone: at cooperatives (certified and non certified) and private trader's level

Table 4.1: Effect of districts (D) and actors (A): certified cooperatives (CC), non-certified cooperatives (NCC) and private traders (PT) on physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty cup quality scores of washed processed coffee beans. Factors PQ PCQ PTQ TSCQ Specific specialty cup quality attributes OCP1 Acidity Body Aroma Flavour AT2 D Mana 25.6 ± 0.6b 48.0 ± 0.7a 73.6 ± 0.8a 82.3 ± 0.8a 7.5 ± 0.2a 7.6 ± 0.1a 7.4 ± 0.1a 7.3 ± 0.1a 7.4 ± 0.1a 7.2 ± 0.1a Goma 28.1 ± 0.8a 47.3 ± 0.8a 75.4 ± 1.4a 81.7 ± 1.3a 7.4 ± 0.3a 7.5 ± 0.2a 7.4 ± 0.2a 7.3 ± 0.2a 7.4 ± 0.2a 7.2 ± 0.2a Limu 27.8 ± 0.8a 49.2 ± 0.7a 77.0 ± 0.7a 80.7 ± 1.1a 7.5 ± 0.2a 7.8 ± 0.3a 7.4 ± 0.2a 7.4 ± 0.1a 7.4 ± 0.1a 7.3 ± 0.1a P- value 0.048 0.16 0.06 0.969 0.939 0.657 0.73 0.73 0.99 0.97 A CC 27.0 ± 0.5a 47.3 ± 0.6b 74.3 ± 1.0b 82.3 ± 0.5a 7.5 ± 0.2b 7.7 ± 0.1a 7.4 ± 0.1a 7.4 ±<0.1a 7.4 ± 0.1a 7.3 ± 0.1a NCC 27.7 ± 0.9a 50.0 ± 0.6a 77.7 ± 1.1a 84.3 ± 0.9a 7.9 ± 0.2a 8.1 ± 0.2a 7.6 ± 0.2a 7.7 ± 0.1a 7.8 ± 0.2a 7.5 ± 0.2a PT 26.8 ± 0.9a 46.8 ± 0.5b 73.6 ± 0.9b 78.8 ± 0.9b 6.9 ± 0.2b 7.2 ± 0.1b 6.9 ± 0.1b 6.9 ±<0.1b 7.1 ± 0.1b 6.9 ± 0.1b P- value 0.54 0.0017 0.009 0.0036 0.027 0.03 0.011 0.003 0.013 0.013 1Overall cup preference, 2After taste, different letters in the same column indicate significant difference, according to Tukey’s HSD post hoc test (P<0.01); are shown as mean ± standard error

Chapter 4

Experiment 2: Coffee quality analysis along value chain in Mana, Goma and Limu districts of Jimma zone: at farmers level

Table 4.2: The interactive effect of actors (A): farmers who are a member of certified cooperatives (FMCC), farmers who are a member of non- certified cooperatives (FMNCC), farmers who are non-member of cooperatives (FNMC) and processing methods (PM): dry (DP) and washed (W) method on physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty cup quality scores A PM PQ PCQ PTQ TSCQ Specific specialty cup quality attributes OCP 1 Acidity Body Aroma Flavour AT2 FMCC DP 37.5 ± 0.8a 48.5 ± 1.1a 86.0 ± 0.6a 84.2 ± 0.3a 7.8 ± 0.1a 7.9 ±<0.1a 7.9 ± 0.1a 7.6 ±<0.1a 7.7± 0.1a 7.7 ±<0.1a W 32.6 ± 1.1a 48.2 ± 1.3a 80.7 ± 0.5a 82.8 ± 0.4b 7.5 ± 0.1b 7.7 ± 0.1b 7.5 ± 0.1a 7.5 ±<0.1a 7.6 ± 0.1a 7.4 ± 0.1a FMNCC DP 37.5 ± 0.6a 50.0 ± 0.7a 87.5 ± 0.8a 84.7 ± 0.3a 7.9 ± 0.1a 7.9 ± 0.1a 7.9 ± 0.1a 7.7 ± 0.1a 7.8 ± 0.1a 7.8 ±<0.1a W 32.6 ± 1.3a 49.0 ± 1.5a 81.6 ± 0.9a 82.9 ± 0.6b 7.7 ± 0.1ab 7.8 ± 0.1b 7.5 ± 0.1a 7.6 ± 0.1a 7.7 ±0.1a 7.6 ± 0.1a FNMC DP 36.3 ± 0.9a 48.3 ±1.3a 84.7 ± 0.7a 83.5 ± 0.3a 7.6 ±<0.1ab 7.9 ±<0.1ab 7.8±<0.1a 7.5 ±<0.1 7.7 ±<0.1a 7.6 ±<0.1a W 32.3 ± 1.0a 48.0 ±1.3a 80.3 ± 0.7a 83.3 ± 0.2a 7.7 ± 0.1ab 7.8 ±<0.1ab 7.6 ±<0.1a 7.5 ±<0.1a 7.6 ±<0.1a 7.5 ±<0.1a P -value 0.784 0.81 0.774 0.038 0.0062 0.039 0.115 0.787 0.76 0.32 1Overall cup preference, 2After taste, different letters in the same column indicate significant difference, according to Tukey’s HSD post hoc test (P<0.01); are shown as mean ± standard error

Analysis of coffee quality along the value chain

Table 4.3: Main effect of district (D), actors (A): farmers who are a member of certified cooperatives (FMCC), farmers who are a member of non-certified cooperatives (FMNCC), farmers who are not member of cooperatives (FNMC) and processing methods (PM): dry (DP) and washed (W) methods on physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty cup quality scores Specific specialty cup quality attributes Factor PQ PCQ PTQ TSCQ OCP 1 Acidity Body Aroma Flavour AT2 D Mana 35.1 ± 0.7a 47.5 ± 0.8b 82.6 ± 0.4a 82.8 ± 0.3b 7.6 ±<0.1b 7.8 ± 0.1a 7.6 ± 0.1b 7.5 ±<0.1a 7.6 ±<0.1b 7.5 ±<0.1b Goma 34.2 ± 0.7a 48.9 ± 0.9ab 83.6 ± 0.5a 83.8 ± 0.3ab 7.8 ± 0.1a 7.9 ±<0.1a 7.8 ±<0.1a 7.5 ±<0.1a 7.7 ±<0.1ab 7.6 ±<0.1ab Limu 34.3 ±1.2a 49.6 ± 1.4a 83.9 ± 0.5a 84.1 ± 0.2a 7.8 ±<0.1a 7.9 ±<0.1a 7.7 ±< 0.1a 7.6 ±<0.1a 7.8 ± 0.1a 7.9 ±<0.1a P -value 0.841 0.0072 0.466 0.008 0.014 0.058 0.014 0.106 0.006 0.031 A FMCC 35.0 ± 0.8a 48.3 ± 0.9b 83.4 ± 0.4a 83.5 ± 0.3a 7.7 ± 0.1a 7.8 ±<0.1a 7.7 ±<0.1a 7.5 ±<0.1a 7.7 ± 0.1a 7.6 ±<0.1a FMNCC 35.0 ± 0.7a 49.5 ± 0.8a 84.5 ± 0.5a 83.8 ± 0.3a 7.8 ± 0.1a 7.9 ± 0.1a 7. 7± 0.1a 7.6 ±<0.1a 7.8 ± 0.1a 7.7 ±<0.1a FNMC 34.3 ± 0.8a 48.2 ± 0.9b 82.5 ±0.5a 83.4 ± 0.2a 7.7 ±<0.1a 7.8 ± <0.1a 7.7 ±<0.1a 7.5 ±<0.1a 7.7 ± <0.1a 7.6 ±<0.1a P -value 0.831 0.04 0.301 0.38 0.31 0.768 0.94 0.0.06 0.31 0.24 PM DP 37.1 ± 0.5a 48.8 ± 0.7a 85.9 ± 0.4a 84.1 ± 0.2a 7.8 ±<0.1a 7.9 ±<0.1a 7.8 ±<0.1a 7.9 ±<0.1a 7.7 ± <0.1a 7.7 ±<0.1a W 32.5 ± 0.6b 48.3 ± 0.8a 80.8 ± 0.4b 82.9 ± 0.2b 7.5 ± 0.1b 7.8 ±<0.1b 7.6± <0.1b 7.5 ±<0.1b 7.5 ± <0.1b 7.4 ±<0.1b P -value <0.0001 0.22 <0.001 <0.001 <0.001 <0.001 <0.001 0.008 0.01 <0.001 1Overall cup preference, 2After taste, different letters in the same column indicate significant difference, according to Tukey’s HSD post hoc test (P<0.01); results are shown as mean ± standard error.

Chapter 4

Experiment 3: Impact of sorting on coffee quality at farmer’s level in Goma district, Jimma zone

Table 4.4: Effect of actors (A): farmers who are a member of certified cooperatives (FMCC), farmers who are a member of non-certified cooperatives (FMNCC), farmers who are non-member of cooperatives (FNMC), sorting treatment (Tr): sorted and unsorted and processing methods (PM): dry (DP) and washed (W) methods on physical quality (PQ), preliminary cup quality (PCQ), preliminary total quality (PTQ), total specialty cup quality (TSCQ) and specific specialty cup quality scores Specific specialty cup quality attributes Factor PQ PCQ PTQ TSCQ OCP 1 Acidity Body Aroma Flavour AT2 A FMCC 38.2 ± 1.2a 44.8 ± 0.6a 83.0 ± 0.9a 83.8 ± 0.7a 7.7 ± 0.1a 7.7 ± 0.1a 7.7 ± 0.1a 7.5 ± 0.1a 7.6 ± 0.1a 7.6 ± 0.1a FMNCC 37.3 ± 0.9a 46.8 ± 0.4a 84.1 ± 0.6a 84.6 ± 0.4a 7.8 ± 0.1a 7.9 ± 0.1a 7.8 ± 0.1a 7.7 ± 0.1a 7.8 ± 0.1a 7.6 ± 0.1a FNMC 38.2 ±0.5 a 44.3 ± 0.4a 82.5 ± 0.5a 83.9 ± 0.5a 7.6 ± 0.1a 7.8 ± 0.1a 7.7 ± 0.1a 7.6 ± 0.1a 7.7 ± 0.1a 7.5 ± 0.1a P -value 0.03 0.79 0.025 0.1 0.13 0.03 0.2 0.36 0.09 0.49 Tr Sorted 38.3 ± 0.7a 47.0 ± 0.3a 85.3 ± 0.6a 85.2 ± 0.3a 7.9 ± 0.1a 7.9 ± 0.1a 7.9 ± 0.1a 7.8 ± 0.1a 7.8 ± 0.1a 7.7 ± 0.1a Unsorted 36.5 ± 0.6b 43.5 ± 0.4a 80.0 ± 0.4b 83.0 ± 0.4 b 7.5 ± 0.1b 7.7 ± 0.1b 7.5 ± 0.1b 7.5 ± 0.1a 7.6 ± 0.1b 7.4 ± 0.1b P 0.004 0.23 0.0004 <0.0001 0.001 0.002 0.003 0.01 0.0028 0.003 PM DP 38.9 ± 0.8a 46.2 ± 0.4a 85.1 ± 0.6a 84.8 ± 0.4a 7.9 ± 0.1a 7.9 ± 0.1a 7.9 ± 0.1a 7.7 ± 0.1a 7.8 ± 0.1a 7.7 ± 0.1a W 35.9 ± 0.7a 45.3 ± 0.4 81.2 ± 0.5b 83.4 ± 0.4b 7.6 ± 0.1b 7.7 ± 0.1b 7.6 ± 0.1a 7.6 ± 0.1a 7.6 ± 0.1a 7.5 ± 0.1a P -value 0.009 0.45 0.0003 0.0009 <0.001 0.002 0.04 0.28 0.02 0.1 1Overall cup preference, 2After taste, different letters in the same column indicate significant difference, according to Tukey’s HSD post hoc test (P<0.01); results are shown as mean ± standard error; results are shown as mean ± standard error.

4.4 Discussion

In Ethiopia a large percentage of coffee export to international market originates from smallholder coffee producers. Coffee produced by smallholder producers, however, reaches the final consumers after passing a number of value chain actors. Unions, cooperatives and private traders are among the value chain actors participating in Ethiopian coffee production and marketing systems. Cooperatives and unions are newly introduced actors with the aim of promoting, modernizing and commercializing smallholder coffee production (Bernard & Spielman, 2009). To enhance market integration and improve livelihoods of smallholder coffee farmers (Stellmacher, 2011), Ethiopian coffee cooperatives are also involved in certification programs. To this effect, some studies (e.g. Jena et al. 2012; Woubie et al., 2015) assessed impacts of coffee certification on livelihoods of Ethiopian smallholder coffee producers. However, there were no studies that evaluated impacts of coffee cooperatives and certifications on green bean quality.

Our study showed that coffee bean quality attributes: preliminary cup quality, preliminary total quality, total specialty cup quality, acidity and aroma showed significant differences among the different coffee actors in the value chain. Coffee beans that passed through a cooperative system had higher quality scores than beans that passed via private traders. This study showed that coffee certification did not improve coffee quality. This is likely so because certification mainly focus on promotion of socio-economic change and environmental sustainability (e.g. organic production ) in coffee production areas (Woubie et al., 2015). Other studies reported as well that certification does not necessarily affect coffee quality but marketability and prices (Bart et al., 2014). Certification also did not bring significant changes on farmer’s income and livelihood (Jena et al., 2012; Mendez et al., 2010; Woubie et al., 2015). This might be due to lack of awareness about certification systems and its procedure.

Since the demand for certified compared to the non-certified coffee beans is stronger (Dempsey, 2006), it is paramount important to further promote the certification schemes along with coffee quality improvement to create awareness at a coffee producer level. The

Chapter 4

certification schemes should also include criteria that improve farmer’s management practice and their livelihood. The certification cost is, however, very high (ca. $6000) per cooperatives and certifications. This probably restricts the practicability of the scheme.

Coffee beans from both certified and non-certified cooperatives were qualified as specialty 1 (Q1) and specialty 2 (Q2) coffees for 33 and 67%, respectively. About 22% of coffee beans from private traders were classified as Q2, while the remaining 78% was classified as commercial coffee grade (grade 3) (Table 4.5).

Table 4.5: Percentage of coffee samples fall in specialty 1 (Q1), specialty 2 (Q2) and grade 3 obtained from different actors: certified cooperatives (CC), non-certified cooperatives (NCC) and private traders (PT). Actors Q1 Q2 Grade 3 CC 33 67 0 NCC 50 50 0 PT 0 22 78

Coffee cooperatives have their own rules on harvesting and handling of coffee beans. They teach and follow up each farmer in the cooperative, for example, on how to harvest and handle coffee cherries. Accordingly, farmers in the cooperatives supply red ripe cherries free from any other foreign materials. Premium prices that are paid by the cooperatives also encourage farmers to give due attention to coffee bean quality. In other studies, (e.g. Kodama (2007) it was also reported that because of premium prices farmers pay more attention for coffees supplied to cooperatives than to private traders. Compared to private traders, coffee cooperatives process coffee beans in their own production sites in such a way that beans harvested from different production sites are less likely mixed with each other.

Farmers-cooperative approach maintains bean qualities, allows traceability and linkages of beans with site of production. A study by Willem (2011) also showed that coffees that pass through cooperative systems had a more specific geographical designation, down to the district and administration level where coffee beans are produced. Hence, cooperatives receive better prices at market than private traders.

Analysis of coffee quality along the value chain

Apart from getting better prices, cooperatives farmers receive input services, training, improved seeds, fertilizer, oil and sugar with a relatively fair prices. Furthermore, cooperatives are considered as a safety net against the exploitation of private traders or in case of international market price reduction risks. Some cooperatives also purchase coffee beans from the non-cooperative farmers. But, there is no any dividend that will be paid back for these farmers. To get immediate cash payments and better prices, some cooperative farmers also prefer to sell their coffee to private traders. This, however, decrease the share of coffee that these farmers submit to their cooperatives.

The growing demand for specialty coffee market opportunities can therefore be much better exploited by coffee cooperative (Dempsey & Campbell, 2006). This indicates that there is potential for Ethiopian coffee cooperatives to produce high quality coffee beans and increase their share in the international specialty coffee market. However, currently the share of coffee cooperatives is small; ca. 15 - 20% compared to the private traders, ca. 80 - 85%. Thus, increasing cooperatives’ share in coffee export might enhance Ethiopian coffee quality on the international market. This can be achieved through institutional interference e.g. ministry of agriculture to encourage and organize farmers into cooperatives. Private traders could also improve their coffee quality, if they develop an institutional arrangement that controls quality of coffee cherries supplied by farmers. In addition, encouraging farmers by providing premium price also improve the quality of supplied coffee beans. However, according to farmers in the area, different qualities of coffee cherries receive equal payment. As the result farmers are not motivated to invest time and energy to improve farming practices (Gelaw, 2016).

Site-specific sample collection and processing might also reduce coffee bean quality deterioration. However, this requires appropriate processing procedures. Our results in experiment 2 support this hypothesis (Table 4.2 and 4.3). In this experiment, coffee bean quality at the farmer’s level under controlled processing conditions was investigated. The results showed that coffee beans from farmers, non-member of cooperatives gave comparable quality to beans coming from farmers that are a member of cooperatives farmers (Table 4.3). The total percentage of coffee samples that were classified as Q1 for farmers, member of certified cooperatives (FMCC), member of non-certified cooperatives 91

Chapter 4

(FMNCC) and non-member of cooperatives (FNMC) for dry processed coffee beans was, respectively 33, 42 and 31% (Table 4.6). This finding, on the other hand, is an “apparent paradox” to the result we observed coffee beans of cooperatives had better quality than private traders. The possible explanation could be the coffee of farmer’s is site specific or less mix compared to the coffee beans of the private traders. The other reason might be the differences in processing conditions e.g. coffee beans of farmers processed under controlled processing conditions. Hence, from these result it can be concluded that the major cause for quality loss by private traders (experiment 1) is due to mixing of coffee beans from different areas, environments (e.g. altitude and shading) and processing methods. Moreover, result of table 4.3 showed slight variation in coffee quality among the districts. This small difference could be due to different in growing conditions.

The result of experiment 2 also revealed that differences processing methods significantly influenced coffee quality attributes e.g. physical quality; preliminary total quality, total specialty cup quality, overall cup preference and acidity. For these quality attributes dry processing method gave higher quality scores than washed methods. Furthermore, the largest percentage of coffee samples classified as Q1 came from dry compared to washed processing method (Table 4.6 and 4.7. However, in Ethiopian, particularly, Jimma zone dry processed coffee beans are often viewed as an inferior quality coffee e.g. DJIMMA GR 5, receiving the lowest price, on average $2.3 per kilogram of green coffee beans. In general the top specialty coffees in the country are usually produced via washed than dry processing method. But the share of coffee bean produced by washed method is small, about 20 - 30% compared to the dry processed coffee bean which accounts 70 - 80%. This may limit the share Ethiopian coffee export on the world speciality coffee market. This indicates that improving dry processed coffee beans can be a better options to make Ethiopian coffee competitive and enhance the share of specialty coffee export. Hence, the present study demonstrated that applying consistent quality control to dry processing can produce high specialty coffee beans e.g. considering ripe red cherries. Moreover, appropriate drying materials also need to be considered to achieve superior quality coffee for dry processing method (Abasanbi, 2010; Subedi, 2011; Tsegaye et al., 2014). These authors reported that coffee cherries dried on raised beds with mesh wire gave higher specialty coffee quality compared to coffee cherries dried on brick floor. This possible due to the raised beds allow 92

Analysis of coffee quality along the value chain better air circulation for the coffee cherries while on the floor the cherries might be in contact with soil or other foreign dirty things that leads to earthy flavour in coffee beverage.

Table 4.6: Percentage of coffee samples fall in specialty 1 (Q1), specialty 2 (Q2) and grade 3 obtained from different actors: farmers, member of certified cooperatives (FMCC), farmers, member of non-certified cooperatives (FMNCC), farmers, non- member of cooperatives (FNMC) and processed by dry and washed methods. Actors PM Q1 Q2 Grade 3 FMCC DP 33 61 6 W 16 78 6 FMNCC DP 42 50 8 W 25 67 8 FNMC DP 31 63 6 W 15 79 6

Thus, from these result we can conclude that main quality differnces between dry processed and washed Ethiopan coffee beans are the different type of coffee cherries being used prior to processing. In actual practice, only red ripe coffee cherries are used for washed processing method while a mixture coffee cherries e.g. unripe, ripe and overripe are usually used for dry processing method. This could be the possible reason for major quality in dry processed coffee beans. In general the study highlighted that dry processing method could also imporove coffee bean quality and increases the percentage of coffee samples that qualify for speciality coffees. Moreover, dry processing is also simpler, cheaper, high throughput and preserves other ecosystem services more than washed processing methods. For instance, washed processing consume huge amount of water, 20 litre per kg coffee cherries (Mburu et al., 1994; von Enden et al., 2002) and has low capacity e.g. 3000 kg h-1.

In general, coffee actors still can improve their coffee quality to produce more specialty coffee, Q1 coffees via sorting and specific site processing. In this study, we also found that sorting coffee cherries before processing improved percentage of Q1 coffee beans from 17 to 61% (Table 4.7). This also indicates that presence of small portion of low quality cherries e.g. unripe and overripe in a certain coffee batch can cause a dramatic quality loss. However, sorting after harvest may leads to quantity reduction. For these selective picking of ripe red cherries and consequently organizing more harvesting rounds for unripe cherries could enhance both quality and quantity of the coffee beans. This might not be controlled all 93

Chapter 4

the time because the cherry collectors use the strip harvesting to increase the amount of coffee cherries as they are paid per volume of the harvested coffee cherries. For these sorting after harvest is mandatory and producers can also keep separately the lowest coffee cherries any way for local sell or consumption. Moreover, farmers have to be trained how inappropriate coffee cherries types can cause a dramatic quality deterioration. Moreover, both cooperatives and private should strictly assess the quality of coffee cherries at the local market.

Table 4.7: Percentage of coffee samples within specialty 1 (Q1) and specialty 2 (Q2) obtained from different actors of Goma district, Jimma zone: farmers who are a member of certified cooperatives (FMCC), farmers who are a member of non-certified cooperatives (FMNCC), farmers who are not member of cooperatives (FNMC), treated (*sorted and **unsorted cherries) and processed by dry and washed methods. Percentage of specialty 1 and 2 Treatment Actors PM Specialty (Q1) Specialty (Q2) Sorted FMCC DP 100 0 WP 33 67 FMNCC DP 100 0 WP 33 67 FNMC DP 100 0 WP 0 100 Mean 61 39 Unsorted FMCC DP 0 100 WP 0 100 FMNCC DP 33 67 WP 0 100 FNMC DP 33 67 WP 33 67 Mean 17 83 *Coffee samples obtained from farmers and further treatment applied (e.g. sorting of unripe cherries, overripe cherries) and **coffee samples obtained from farmers and processed without further treatment/ sorting.

4.5 Conclusions

The study in general showed that coffee beans managed by cooperatives had better quality scores than beans managed by private traders. Coffee certification, on the other hand, did not allow any quality improvement. Coffee beans from farmers, non- member of cooperatives had better quality than coffee beans of private traders. The study also revealed that dry processing method improved coffee bean quality and the percentage of these beans

Analysis of coffee quality along the value chain falling in Q1 grades than washed processed coffee beans. However, to enhance coffee quality of dry processing method, proper coffee cherries e.g. ripe red and clean cherries needed to be considered. Furthermore, dry processing method conserves the environment over washed method. Sorting of unripe and overripe coffee cherries also improved coffee quality and all sorted coffee samples of dry processed coffee beans falls in Q1 coffee. Further studies that consider different seasons and other coffee growing regions can substantiate this conclusion.

Chapter 5

Prediction of specialty coffee cup quality based on near infrared spectra of green coffee beans

Total specialty cup quality

2 Green beans NIR-machine RMSECV: 1.31; R cv : 81 Spectra 90 RMSEP: 1.04; r: 90

85 ed t c i ed

r 80 Predicted

70 Specialty cup quality Specialty coffee categories attributes and points 70 75 80 85 90 95 Grade Total points MeasuredMeasur Cup cleanness (10) Specialty1(Q1) 85-100 Acidity (10) Specialty 2(Q2) 80-84.75 Body (10) 71-80 Flavor (10) Grade 3 Aroma (10) After taste (10) Uniformity (10) Sweetness (10) Balance (10) Overall cup preference (10) Tot al specialty cup quality (100) Roasted beans Cup tasting Coffee quality grading

This chapter has been publihsed as

Tolessa, K., Rademaker, M., De Baets, B., and Boeckx, P. (2016). Prediction of specialty coffee cup quality based on near infrared spectra of green coffee beans. Talanta, 150, 367-374.

Chapter 5

Abstract

The growing global demand for specialty coffee increases the need for improved coffee quality assessment methods. Green bean coffee quality analysis is usually carried out by physical (e.g. black beans, immature beans) and cup quality (e.g. acidity, flavor) evaluation. However, these evaluation methods are subjective, costly, time consuming, require sample preparation and may end up in poor grading systems. This calls for the development of a rapid, low-cost, reliable and reproducible analytical method to evaluate coffee quality attributes and eventually chemical compounds of interest (e.g. chlorogenic acid) in coffee beans. The aim of this study was to develop a model able to predict coffee cup quality based on NIR spectra of green coffee beans. NIR spectra of 86 samples of green Arabica beans of varying quality were analyzed. Partial least squares (PLS) regression method was used to develop a model correlating spectral data to cupping score data (cup quality). The selected PLS model had a good predictive power for total specialty cup quality and its individual quality attributes (overall cup preference, acidity, body and aftertaste) showing a high correlation coefficient with r-values of 90, 90,78, 72 and 72, respectively, between measured and predicted cupping scores for 20 out of 86 samples. The corresponding root mean square error of prediction (RMSEP) was 1.04, 0.22, 0.27, 0.24 and 0.27 for total specialty cup quality, overall cup preference, acidity, body and aftertaste, respectively. The results obtained suggest that NIR spectra of green coffee beans are a promising tool for fast and accurate prediction of coffee quality and for classifying green coffee beans into different specialty grades. However, the model should be further tested for coffee samples from different regions in Ethiopia and test if one generic or region-specific model should be developed.

Key words: Coffee quality, NIR spectra, PLS model and Specialty coffee

Prediction of specialty coffee cup quality

5.1 Introduction

The specialty coffee market is constantly growing, which is driven by changing consumers preferences (Oberthür et al., 2012). Specialty coffee refers to the highest quality of green coffee beans, has a known geographical origin, and mostly includes coffee beans certified as e.g. organic coffee (coffee grown without chemical fertilizer and pesticides), fair trade (promotes greater economic incentives via fair price and direct trade) and rainforest alliance (integrates biodiversity conservations, community development and suitable agricultural practices to ensure sustainable farm management) (Adam, 2009). High quality green coffee beans show little or no physical defects (e.g. black beans, immature beans, floaters, broken beans, insect damage and other foreign materials such as seeds of shade trees) and, when roasted, have a distinctive character in the cup and high cup tasting scores (Mateos-Aparicio, 2011; Rodriguez-Saona & Allendorf, 2011). As such, the success of specialty coffee brings along a need for more reliable and objective quality assessment methods.

The increasing consumption of specialty coffee (Oberthür et al., 2012) creates a strong potential and opportunity for coffee-producing countries. An increasing demand for such coffees on the world market resulted in a segmentation of the coffee market and higher prices are paid for specialty coffees (Labouisse et al., 2008). For instance, specialty coffee beans receive a premium price of ca. 20-50% compared to regular coffee beans. Certification also raises coffee prices (e.g. organic and fair trade generate premium of about ca. 9%) (Bart et al., 2014). Ethiopia is one of the top ten coffee-producing countries in the world and the largest exporter in Africa (Bart et al., 2014). The country produces specialty coffees with a wide range of flavors (ICCO, 2014). Currently, specialty coffee accounts for ca. 20% of Ethiopia’s coffee export and there is potential to boost its share in the world market even further (Robert, 2015).

The assessment of coffee quality is a key step in price setting and determines its export potential in coffee-producing countries. Consequently, it is of major importance to many coffee roasters and distributors (Santos et al., 2012). Currently, physical and cup quality analysis are the key tools for coffee quality assessment. Physical quality of green coffee beans is associated with the presence of defects found in certain coffee batches such as

Chapter 5

deviations in odor, colour, size and shape of beans. Cup quality characteristics are attributes of coffee that can be distinguished by senses and can be assessed organoleptically by professional coffee tasters, based on established terminologies for cup quality analysis (e.g. flavor, acidity and body and cup cleanness). Based on these quality attributes the grading system for coffee varies and follows specific guidelines of the producing country (Santos et al., 2012).

In the Ethiopian system, physical and cup quality analysis enable experts to evaluate total preliminary coffee quality, which is used to classify coffee beans into different grades (see detail in chapter 1; Table 1.1). Coffee beans graded from grade one to three (total preliminary quality > 70%) are grouped into specialty coffee. Coffee beans from these three grades are further classified into different specialty grades, which are called: Q1, Q2 and commercial (see detail in chapter 1; Table 1.2). However, such methods of assessing coffee quality are rather subjective (Feria-Morales, 2002). In addition, they are costly, require sample preparation and are time consuming. They are also difficult to implement when many samples need to be analyzed during peak coffee production. Therefore, there is a need for a rapid, low-cost, reliable and reproducible analytical method as an alternative tool to evaluate coffee quality attributes and, eventually, other chemical compounds presents in coffee beans.

Near-infrared spectroscopy (NIRS) has gained broad acceptance for food quality evaluation (Rodriguez-Saona & Allendorf, 2011). It is rapid, reliable, simple and cheap, does not require use of chemicals and requires little sample preparation as well as low instrumental maintenance (Jing et al., 2010; Krähmer et al., 2015; Santos et al., 2012). Several studies have shown the potential applicability of NIRS as an alternative technique to analyze coffee quality. NIRS has been used for the discrimination of Arabica coffee from Robusta (Buratti et al., 2014; Downey et al., 1997; Esteban-Dıez et al., 2004; Esteban-Díez et al., 2007; Kemsley et al., 1995; Santos et al., 2012), immature from mature as well as defective from non-defective coffee beans (Craig et al., 2011, 2012; Craig et al., 2012; Craig et al., 2015; Santos et al., 2012). It was also reported that NIRS could be used successfully for authentication of coffee origin (Downey & Boussion, 1996; Wang et al., 2009), detection of coffee adulteration (Briandet et al., 1996; Downey &

100

Prediction of specialty coffee cup quality

Spengler, 1996; Ebrahimi-Najafabadi et al., 2012; Reis et al., 2013), quantification of theobromine, theophylline, ash and lipid content of roasted coffee beans (Consuelo et al., 2007; Huck et al., 2005; Pizarro et al., 2004; Zhang et al., 2013). Some other studies also indicated the feasibility of predicting coffees cup quality attributes from roasted coffee beans using NIRS (Ribeiro et al., 2011).

However, NIRS-based coffee cup quality prediction has been mainly focused on roasted beans and only a few studies used green coffee beans for coffee discrimination between defective and non-defective beans and to evaluate physical bean quality (Craig et al., 2012, 2012; Ribeiro et al., 2012). Many countries, organisations and companies could benefit from a reliable NIRS-based coffee quality assessment starting from green beans. However, to the best of our knowledge, prediction of coffee cup quality from NIRS spectra of green coffee bean has not yet been reported. Therefore, the objective of the present study was to develop a NIRS-based model for the prediction of coffee cup quality from green coffee beans using partial least squares regression.

5.2 Material and Methods

5.2.1 Description of the study site . The coffee samples were collected from three districts (Mana, Goma and Limu Kosa) of the Jimma zone, Oromia regional state of Ethiopia. They are known as the predominant Coffea arabica growing areas of the Jimma zone. Sample collection was carried out from October 2012 to December 2013 during the peak harvesting season.

5.2.2 Coffee samples

The data set used in the present study comprised 86 coffee samples with varying cup quality characteristics as the coffee samples were collected from different environments (low, mid and high altitude) (Supplementary information Table 5.1). Only fully ripe, red coloured coffee cherries were harvested by hand picking and then subjected to dry,washed and semi-washed processing methods.

101

Chapter 5

5.2.3 Coffee quality analysis

From each sample, 350 g green beans were delivered to the Ethiopia Commodity Exchange (ECX), Jimma Branch, for physical and cup quality evaluation. A team of three professional who operate commercially in the ECX carried out the cup taste analysis. During the preparation of the actual cup tasting, 100 g green beans of each sample was roasted at 200°C for 8-12 min using a sample roasting machine (Probat, 4 Barrel Roaster, Germany) allowing medium roasting of the coffee. The roasted beans were tipped out into a cooling tray and rapidly cooled by blowing cold air through the beans for four minutes. Then the roasted beans were ground with a coffee grinding machine (Mahlkonig Guatemala SB, K32SB2, Germany) with medium adjustment. For cup quality analysis, 13.75 g of ground coffee was diluted in a cup of 250 ml of hot water (93°C) at the time of pouring onto the ground coffee to prepare an infusion. Five cups of brewed coffee for each coffee sample were prepared for analysis and three cuppers tasted each of the five cups. Subsequently, the average score of each cupper was taken as replication. The preliminary total quality scoring comprises physical (40/100) and cup quality (60/100) scores. The criteria commonly used to evaluate the physical quality of coffee beans include the presence of defects: primary and secondary defects, shape and make, odour and colour. For cup quality assessment, acidity, body, flavor and cup cleanness were considered and the sum of these four cup quality attributes gives a total cup quality with a score between 0-60. Hence, the sum of both cup and physical quality ranges between 0 and 100 and is used to classify the samples into different grades (see detail in chapter 1, section 1.6, Table 1.1). Specialty quality analysis was done based on overall cup preference, acidity, body, aroma, flavor, aftertaste, uniformity, cup cleanness, sweetness and balance each on a scale ranging from 0 to 10 (see detail in chapter 1, section 1.6, Table 1.2). Based on the total specialty coffee scores, samples were further grouped into specialty 1 (Q1) ≥ 85, specialty 2 (Q2) (80-84.75) and grade 3 (71-80) (John, 2011).

102

Prediction of specialty coffee cup quality

5.2.4 NIR spectra measurement and analysis

A Fourier Transform-NIR spectrometer (Tango, Brücker, Belgium) was used to obtain three replicated spectra. Green coffee samples were ground to a size smaller than 5 mm and 15 g of each sample was used for analysis. The samples were irradiated with tungsten (5V/7W) as source of near infrared light. The spectra were recorded in the wavenumber range between 4000 and 11500 cm-1 with a spectral resolution of 8 cm−1 and the reflectance was detected by an Indium Gallium Arsenide (InGaAs) diode. The measurement was performed at room temperature. Pre-processing of original NIR spectra to remove unwanted information, such as spectral noise, was a crucial step prior to model development since it deteriorates the quality of the model (Baptista et al., 2008; Craig et al., 2015). As such, specific pre-processing methods were identified by the OPUS QUANT software for each coffee quality attribute (Supplementary information; Table 5.2). Thus, suitable processing methods were selected based on the lowest root mean square error of cross validation (RMSECV). Derivatives and vector normalisation methods were used for preliminary cup quality, flavour and aftertaste, subtractions of straight line for preliminary total quality, multiplicative scatter correction for total specialty cup quality, first derivatives and multiplicative scatter correction for overall cup preference and body, vector normalisation for acidity and first derivatives for aroma.

To establish a linear relationship between spectral data and cupping scores, Partial Least Squares (PLS) regression analysis was performed (Ferreira et al., 1999; Ribeiro et al., 2009) with OPUS QUANT software version 7.2 (Brücker, Belgium). The coffee samples were divided into calibration (n = 66) and prediction (n = 20) samples (Supplementary information; Table 5.2). The calibration samples were randomly selected and used for model development using the leave-one-out cross-validation procedure. The procedure was repeated untill all calibration data sets were left one by one. Prediction samples (i.e. samples not used for model development) were used for external validation of the developed model. The optimum number of factors were determined on the basis of the lowest RMSECV value (Baptista et al., 2008). The performance of the developed PLS model was evaluated in terms of RMSECV, root mean square error of external validation (prediction) (RMSEP), residual prediction 2 deviation (RPD), coefficient of determination for cross validation (R cv) and correlation

103

Chapter 5

coefficient (r) for external validation (Baptista et al., 2008; Craig et al., 2015; Xie et al., 2009). For a good model, the value of RMSECV and RMSEP needs to be as low as possible and gives high R2 and RPD values. The value of RMSECV and RMSEP refers to the quantitative measure for the precision of the predicted samples in cross and external validation, respectively.

5.3 Results

Means and ranges of measured cupping scores of all coffee quality attributes for cross- validation and prediction samples are shown in Table 5.5. Quality analysis (both preliminary and specialty cup quality) R2cv, RMSECV and RPD values of the cross validation data set ranged between 48 – 89, 0.19 – 3.63 and 1.39 – 3.02, respectively. In the prediction data set the corresponding values were 0.01 – 90, 0.22 – 4.82 and 0.74 – 2.25, respectively for r, RMSEP and RPD (Table 5.5). Preliminary total quality had lower R2cv (48) compared to preliminary cup quality (R2cv = 67) and total specialty cup quality (R2cv = 81) (Table 5.3, Fig.5.1). For the specialty cup quality attributes, the highest R2cv was found for overall cup preference (R2cv = 89), followed by acidity (R2cv = 83), total specialty cup quality (R2cv = 81), aftertaste (R2cv = 80) and body (R2cv = 78). Flavour (R2cv = 66) and aroma (R2cv = 59), on the other hand, had the lowest R2cv values (Table 5.3, Fig.5.2).

The RMSECV values for preliminary cup quality, preliminary total quality, total specialty cup quality, overall cup preference, acidity, body, aroma, flavour and aftertaste were 1.73, 3.63,1.31, 0.19, 0.21, 0.22, 0.33,0.29 and 0.22, respectively (Figs. 5.1 and 5.2). The highest RPD value of a cross-validation data set, was obtained for overall cup preference (3.02), followed by acidity (2.44), total specialty cup quality (2.27), aftertaste (2.25), body (2.13). Preliminary cup quality (1.73), flavour (1.71) aroma (1.57) and total preliminary quality (1.39) relatively had lowest RPD values (Table 5.3, Fig.5.1).

For the prediction data set, the PLS models for overall cup preference and total specialty cup quality had the highest coefficient of correlation, (r = 90), followed by acidity (78), aftertaste (72) and body (72). Preliminary cup quality (27) and aroma (0.01), on the other

104

Prediction of specialty coffee cup quality hand, had the lowest coefficient of correlation. The PLS model of total specialty cup quality and that of individual specialty coffee quality attributes had the lowest RMSEP values compared to preliminary total and preliminary cup quality PLS models. The highest RMSEP value (4.82) was found in preliminary total quality while the lowest value (0.22) was observed in overall cup preference. In the same data set, overall cup preference had the highest RPD (2.25) while the lowest values were observed in aroma and aftertaste (Fig.5.2).

105

Chapter 5

Table 5.3: Statistics for cross validation and prediction (external validation) data for preliminary and specialty coffee quality Coffee quality attributes Cross validation data set Prediction data set

2 Number of Mean Range R cv RMSECV RPD Mean Range r RMSEP RPD factors

Preliminary cup quality 8 48.4 42-54 67 1.72 1.73 49.2 42-57 27 3.51 1.01 Preliminary total quality 4 82.2 71-91 48 3.63 1.39 82.9 71-91 59 4.82 1.24 Total specialty cup quality 9 84.7 76.8 - 92.5 81 1.31 2.27 84.2 79.5 - 87.5 90 1.04 1.98 Specific specialty cup quality

attributes Overall cup preference 10 7.9 6.5 - 9.0 89 0.19 3.02 7.9 7.0 - 9.0 90 0.22 2.25 Acidity 4 8.0 6.75 - 9.0 83 0.21 2.44 7.9 7.0 - 9.0 78 0.27 1.56 Body 8 7.9 6.5 - 9.0 78 0.22 2.13 7.7 7.3 - 8.3 72 0.24 1.12 Aroma 3 7.8 6.5 - 9.0 59 0.33 1.57 7.6 7.0 - 8.0 0.01 0.40 0.74 Flavour 10 7.8 6.5 - 9.0 66 0.29 1.71 7.8 7.3 - 8.3 59 0.29 1.01 Aftertaste 10 7.7 6.5 - 9.0 80 0.22 2.25 7.7 7.0 - 8.0 72 0.27 0.96 2 RMSECV: Root Mean Square Error of Cross Validation; RMSEP: Root Mean Square Error of Prediction; RPD: Residual Predictive Deviation; R cv: coefficient of determination in cross validation and r: coefficient of correlation in prediction data set

106

Prediction of specialty coffee cup quality

60 95 95 2 2 RMSECV: 3.63; R cv: 48 RMSECV: 1.72; Rcv: 67 RMSECV: 1.31; R 2 : 81 RMSEP: 4.82; r: 59 cv RMSEP: 3.51; r: 27 56 90 90 RMSEP: 1.04; r: 90

85 85 52

48 80 80 Predicted Predicted Predicted

44 75 75

40 70 70 40 44 48 52 56 60 70 75 80 85 90 95 70 75 80 85 90 95 Measured Measured

(a) Preliminary cup quality (b) Preliminary total quality (c) Total specialty cup quality

Fig.5.1: Measured vs. predicted scores of preliminary cup quality (a), preliminary total quality (b) and total specialty cup quality (c); data set for cross validation (empty bullets; dashed line) and for predictions (full bullets, full line). RMSECV: Root Mean Square Error of Cross Validation; RMSEP: Root 2 Mean Square Error of Prediction, R cv: coefficient of determination in cross validation and r: coefficient of correlation for prediction data set.

107

Chapter 5

9.5 9.5 9.5 2 RMSECV: 0.19; R cv: 89 2 RMSECV: 0.21; R2 : 83 RMSECV: 0.22; R cv: 78 9.0 RMSEP: 0.22; r: 90 9.0 cv 9.0 RMSEP: 0.27; r: 78 RMSEP: 0.24; r: 72 8.5 8.5 8.5

8.0 8.0 8.0

7.5 Predicted 7.5 Predicted 7.5 Predicted

7.0 7.0 7.0

6.5 6.5 6.5 c a b 6.0 6.0 6.0 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 Measured Measured Measured

(a) Overall cup preference (b) Acidity (c) Body 9.5 9.5 9.5 2 2 2 RMSECV: 0.29; R cv : 66 RMSECV: 0.22; R cv : 80 RMSECV: 0.33; R cv: 59 9.0 9.0 RMSEP: 0.29; r: 58 RMSEP: 0.27; r: 72 9.0 RMSEP: 0.40; r: 0.01 8.5 8.5

8.5

8.0 8.0 8.0

Predicted 7.5

7.5 Predicted Predicted 7.5 7.0 7.0 7.0 d e 6.5 f 6.5 6.5 6.0 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 Measured Measured Measured (d) Aroma (e) Flavour (f) Aftertaste Fig.5.2: Measured vs. predicted scores of overall cup preference (a), acidity (b), body (c), aroma (d), flavour (e) and aftertaste (f); data set for cross validation (empty bullets; dashed line) and for predictions (full bullets, full line). RMSECV: Root Mean Square Error of Cross Validation; Root Mean Square Error of Prediction, R2cv: coefficient of determination in cross-validation and r: coefficient of correlation in prediction data set

108

Prediction of specialty coffee cup quality

5.4 Discussion

Different PLS models were developed on the basis of NIR spectra of green coffee beans to predict several coffee quality properties. Very promising results were obtained for total specialty cup quality and its individual quality attributes (overall cup preference, acidity, body, flavour and aftertaste) compared to results obtained for preliminary quality. The PLS models of total specialty cup quality, overall cup preference, acidity, body, flavour and aftertaste had, in general, higher R2cv, lower RMSECV and higher RPD values compared to the corresponding values of preliminary cup quality and preliminary total quality.

The correlation between measured and predicted cupping scores for total specialty cup quality (r = 90), overall cup preferences (r = 90), acidity, body and aftertaste were also high, which was, not the case for total preliminary quality (r = 59) (Figs. 5.1b and 5.1c). This lower prediction efficiency ("closeness" of the model result to the measured value) for total preliminary quality was most likely due to physical quality parameters (e.g. defects, odour and colour), which are possibly more difficult to detect via NIRS (Barbin et al., 2014). Thought the defects could not be recognised individually (Barbin et al., 2014) indicated that NIRS is still a potential tool to enable fast, simple and quantitative evaluations of green coffees. The higher RPD and the lower RMSEP for total and individual specialty cup quality attributes compared to preliminary quality also showed the promising predictive capabilities of the developed models. Aroma, on the other hand, had higher RMSECV, RMSEP and lower R2cv values compared to other specialty quality attributes. This indicates that the model had far less power to predict the aroma content, probably due to the complex nature of the attribute and because Coffea arabica contains over 40 major aroma components (Aeschbacher et al., 1989). Further complicating aroma content prediction is the fact that the aroma is a result of a number of chemical reactions that occur during roasting (Fisk et al., 2012).

Numerous efforts have already been carried to use NIRS as alternative technique to determine coffee quality attributes and to analyse its chemical composition and also reported that the techniques is even better than other analytical techniques such as HPLC (e.g. chlorogenic acid and caffeine) (De Maria et al., 1994). Ribeiro et al. (2011),

109

Chapter 5

indicated that NIRS can be used to establish the relationship between coffee cup quality attributes and the chemical composition of roasted coffee beans. The authors also developed a model from roasted beans to predict coffee cup quality attributes (e.g. overall cup preference, acidity, body, flavour and cleanliness). Our study developed a model from green beans with a higher R2cv and a lower RMSEP value for all cup quality attributes (overall cup preference, acidity and body) except for flavour, which had a lower R2cv. The reason for such discrepancy might be due to the number of samples used for model development and external validation. The number of samples has an impact on the quality of the model developed (e.g. the higher the number of samples the better predictive power of the model). In our study a total of 86 coffee samples were used, while (Ribeiro et al., 2011) used 51 samples.

The current study also indicates the potentiality of directly using green coffee beans to evaluate coffee quality. Prediction of coffee quality from green coffee beans is faster and cheaper than from roasted coffee beans. Apart from specific quality attributes our study developed promising models for total specialty cup quality, which can be used for classifying coffees into different specialty grades. For instance, out of 11 samples in “specialty 1”, nine of them are correctly predicted and two samples were predicted as “specialty 2” by our developed model. All eight samples found in “specialty 2” were predicted correctly. This showed that NIRS can potentially also be used to distinguish specialty coffees.

5.5 Conclusions

Specialty cup quality could be predicted well by the developed NIR-based prediction model. Prediction of total preliminary quality was relatively poor, perhaps indicating that the physical quality assessment can’t easily be done by NIR spectra of green beans. The NIR-based technique proved promising to predict some individual quality attributes such as overall preference, acidity and body. However, there is a need for improvement for total preliminary quality, flavour and aroma. Our model for total specialty cup quality holds potential as a rapid and reliable classification method for assigning coffees into different specialty grades. Hence coffee quality analysis with NIR spectra of green coffee

110

Prediction of specialty coffee cup quality beans is promising. However, the developed models need to be tested now further on independent data sets from other Ethiopian regions to test stability and accuracy of the models.

111

Chapter 5

5.6 Supplementary information Table 5.1: Overview of 86 coffee samples used for model development and validation with measured values of quality attributes Total Sample Processing Preliminary Preliminary specialty Overall cup No. identifiers District Altitudea method total quality cup quality cup quality preference b Acidity Body Aroma Flavour Aftertaste 1 741DP Limu Kosa Mid Dry 79.00 45.00 81.50 7.25 7.50 7.25 7.50 7.50 7.25 Semi- 2 7440SW Limu Kosa Mid washed 73.00 48.00 81.75 7.25 7.75 7.50 7.25 7.25 7.50 3 74110DP Limu Kosa Mid Dry 79.00 45.00 81.50 7.25 7.50 7.25 7.25 7.50 7.25 Semi- 4 74148SW Limu Kosa Mid washed 85.00 48.00 82.00 7.50 7.75 7.50 7.25 7.50 7.00 5 74148WP Limu Kosa Mid Washed 74.00 45.00 80.25 7.00 7.25 7.25 7.50 7.25 7.00 6 74165DP Limu Kosa Mid Dry 76.00 42.00 80.25 7.00 7.25 7.25 7.25 7.50 7.00 Semi- 7 74165SW Limu Kosa Mid washed 79.00 45.00 81.00 7.25 7.50 7.25 7.25 7.25 7.25 8 74165WP Limu Kosa Mid Washed 74.00 45.00 82.25 7.25 7.75 7.50 7.25 7.50 7.50 9 75227DP Limu Kosa Mid Dry 76.00 42.00 79.50 7.00 7.75 7.25 7.00 7.25 7.00 10 AFATAF1WP Mana Mid Washed 77.00 45.00 80.75 7.25 7.50 7.50 7.25 7.25 7.00 11 CHOCE-C-WP Goma Mid Washed 71.00 45.00 81.75 7.25 7.75 7.50 7.50 7.50 7.25 12 CHOCE-F1WP Goma Mid Washed 82.00 45.00 82.75 7.50 6.75 7.50 7.50 7.50 7.50 13 COMB DP1 Limu Kosa Mid Dry 82.00 45.00 83.50 7.75 8.00 7.75 7.50 7.50 7.50 14 COMB DP2 Limu Kosa Mid Dry 82.00 45.00 81.50 7.25 8.00 7.50 7.25 7.25 7.25 15 COMB DP3 Limu Kosa Mid Dry 79.00 45.00 83.00 7.50 8.00 7.75 7.50 7.50 7.50 16 COMB WP Limu Kosa Mid Washed 73.00 48.00 76.75 6.50 8.00 6.50 6.50 6.50 6.50 17 DAWA-NC Goma High Washed 71.00 45.00 80.25b 7.00 8.00 7.00 7.50 7.25 7.00 18 DEBF1WP Limu Kosa Mid Washed 83.00 51.00 85.25 8.00 7.75 7.75 7.50 8.25 8.00 19 DOYO-NC Goma High Washed 75.00 51.00 86.00 8.50 7.75 8.00 7.50 8.00 7.50 20 HAROF2DP Mana High Dry 88.00 48.00 84.25 7.75 7.75 7.75 7.50 7.75 7.75 21 HF2DP Mana High Dry 88.00 51.00 86.75 8.50 7.75 8.00 8.00 8.25 8.00 Semi- 22 HF2SW Mana High washed 82.00 48.00 85.75 8.00 7.75 7.75 7.75 8.25 8.00

112

Prediction of specialty coffee cup quality

23 HF2WP Mana High Washed 83.00 51.00 85.75 8.00 7.75 8.00 7.75 8.25 7.75 24 HF4DP Mana High Dry 91.00 51.00 86.50 8.50 7.50 8.00 8.00 8.25 7.75 Semi- 25 HF4SW Mana High washed 81.00 45.00 85.00 8.00 7.75 7.75 7.75 8.00 7.75 26 HF4WP Mana High Washed 81.00 48.00 84.25 7.75 7.75 7.75 7.75 8.00 7.50 27 HUDAF1WP Goma High Washed 86.00 51.00 85.00 8.25 7.75 7.75 7.50 8.00 7.75 28 ILBUF1DP Goma Mid Dry 85.00 48.00 84.00 7.75 9.00 8.00 7.50 7.50 7.50 29 KENTERF1WP Mana High Washed 81.00 48.00 80.50 7.25 9.00 7.25 7.25 7.25 7.00 30 LF1DP Mana Low Dry 82.00 45.00 83.50 7.75 9.00 7.50 7.50 7.75 7.75 Semi- 31 LF1SW Mana Low washed 79.00 48.00 83.50 7.50 8.75 7.75 7.50 7.50 7.75 32 LF1WP Mana Low Washed 81.00 48.00 83.25 7.50 8.25 7.75 7.50 7.75 7.50 33 LF3DP Mana Low Dry 82.00 45.00 82.25 7.50 9.00 7.50 7.50 7.50 7.50 Semi- 34 LF3SW Mana Low washed 78.00 48.00 83.25 7.50 8.50 7.75 7.50 7.50 7.50 35 LF3WP Mana Low Washed 75.00 45.00 82.00 7.50 8.00 7.50 7.50 7.50 7.50 Semi- 36 LF5SW Mana Low washed 85.00 48.00 83.75 7.75 9.00 7.75 7.50 7.75 7.50 37 LF5DP Mana Low Dry 79.00 45.00 83.25 7.75 8.25 7.50 7.50 7.50 7.50 38 LF5WP Mana Low Washed 81.00 48.00 83.50 7.50 8.50 7.75 7.50 7.75 7.50 39 LPS1WP Limu Kosa Mid Washed 80.00 48.00 84.00 7.75 8.25 7.50 8.00 7.50 7.75 40 LPS3WP Limu Kosa Mid Washed 75.00 51.00 85.75 8.00 9.00 7.75 8.25 8.00 7.75 Semi- 41 MF1WP Mana Mid washed 84.00 48.00 82.50 7.50 9.00 7.50 7.50 7.50 7.25 42 OMOF1WP Goma Mid Washed 85.00 48.00 86.50 8.25 8.75 8.25 7.75 8.00 8.00 43 PSAF6WP Limu Kosa Mid Washed 75.00 51.00 84.00 7.75 9.00 8.00 7.50 7.75 7.50 44 PSHUF1WP Goma High Washed 88.00 51.00 83.75 8.00 8.75 7.50 7.50 7.75 7.50 45 SHGF2WP Limu Kosa Mid Washed 88.00 51.00 85.75 8.00 7.75 7.75 7.75 8.25 8.00 46 HF1-FH-D Mana High Dry 82.00 45.00 90.25 9.00 8.00 9.00 7.75 9.00 9.00 47 HF1-FH-M Mana High Dry 91.00 54.00 92.25 8.75 8.50 9.00 9.00 9.00 9.00 48 HF1-FH-O Mana High Dry 91.00 54.00 92.00 9.00 7.25 9.00 9.00 9.00 9.00 49 HF1-SH-D Mana High Dry 85.00 51.00 87.00 8.50 7.50 7.75 9.00 7.75 7.50

113

Chapter 5

50 HF1-SH-M Mana High Dry 88.00 48.00 85.50 8.25 8.50 7.75 8.00 7.75 7.75 51 HF1-SH-O Mana High Dry 88.00 54.00 92.00 9.00 8.00 9.00 9.00 9.00 9.00 52 HF1-TH-M Mana High Dry 88.00 51.00 89.50 8.50 7.50 8.50 8.50 8.50 9.00 53 HF1-TH-O Mana High Dry 85.00 48.00 84.00 8.00 7.25 7.75 7.75 7.50 7.50 54 HF2-FH-D Mana High Dry 85.00 48.00 84.75 7.50 8.50 8.25 7.50 7.50 7.50 55 HF2-FH-M Mana High Dry 82.00 48.00 85.25 8.25 8.00 8.25 7.75 7.75 7.50 56 HF2-FH-O Mana High Dry 85.00 51.00 89.00 8.50 8.00 8.50 8.50 8.75 8.25 57 HF2-SH-M Mana High Dry 85.00 48.00 85.50 8.25 8.25 8.25 8.75 7.50 7.50 58 HF3-FH-D Mana High Dry 85.00 51.00 86.50 8.00 8.00 9.00 8.00 8.00 7.75 59 HF3-FH-M Mana High Dry 85.00 51.00 90.25 8.75 8.00 9.00 8.25 9.00 8.50 60 HF3-SH-M Mana High Dry 85.00 48.00 86.25 8.75 8.00 8.75 7.75 7.75 7.75 61 HF3-SH-O Mana High Dry 88.00 54.00 92.50 9.00 8.00 9.00 9.00 9.00 9.00 62 HF3-TH-M Mana High Dry 91.00 54.00 90.25 8.75 7.75 8.75 8.75 8.50 8.75 63 HF4-TH-M Mana High Dry 88.00 51.00 86.25 8.00 7.75 7.75 8.50 7.75 8.50 64 HF5-FH-D Mana High Dry 91.00 51.00 85.50 8.00 7.00 8.00 8.00 8.00 7.75 65 HF5-FH-M Mana High Dry 85.00 51.00 87.50 8.50 7.75 8.50 8.50 8.00 8.50 66 DUR-NC Goma High Washed 78.00 54.00 88.00 8.50 8.00 8.50 7.75 8.50 8.25 67 744WP Limu Kosa Mid Washed 81.00 48.00 82.25 7.25 7.50 7.50 7.25 7.50 7.50 68 74112WP Limu Kosa Mid Washed 71.00 42.00 79.75 7.00 7.00 7.25 7.00 7.50 7.00 69 74158DP Limu Kosa Mid Dry 85.00 48.00 83.25 7.50 8.00 7.75 7.50 7.50 7.50 70 HF1WP Mana High Washed 85.00 51.00 85.50 8.00 8.00 7.75 7.75 8.25 7.75 71 HF5DP Mana High Dry 88.00 51.00 85.50 8.25 8.00 7.75 8.00 8.00 7.75 Semi- 72 HF5SW Mana High washed 84.00 54.00 86.50 8.00 8.00 8.00 8.00 8.25 8.00 73 LF2DP Mana Low Dry 85.00 48.00 83.50 7.75 7.75 7.50 7.50 7.75 7.75 74 LF4DP Mana Low Dry 82.00 45.00 82.75 7.50 7.75 7.50 7.50 7.50 7.50 75 LF4WP Mana Low Washed 78.00 45.00 81.50 7.25 7.50 7.25 7.50 7.50 7.25 76 DURF1DP Goma High Dry 91.00 57.00 87.50 8.25 8.25 8.25 8.00 8.25 8.00 77 OMOF2DP Goma High Dry 91.00 51.00 85.75 8.00 8.00 8.00 7.75 8.00 8.00

114

Prediction of specialty coffee cup quality

78 DEBF2WP Limu Kosa Mid Washed 87.00 51.00 85.00 8.00 8.00 8.00 7.50 8.00 7.75 79 HUNDA-NC-WP Goma High Washed 76.00 48.00 84.75 8.50 8.00 7.50 8.00 7.75 7.50 80 MPS1WP Mana Mid Washed 74.00 51.00 85.00 8.00 8.00 8.00 8.00 7.25 7.75 81 HF5WP Mana High Washed 83.00 51.00 85.00 8.00 8.00 7.75 7.75 8.00 7.75 82 GPS3 Goma Mid Washed 85.00 51.00 84.75 8.00 7.75 8.00 8.00 7.50 7.75 83 HF5-FH-O Mana High Dry 85.00 51.00 87.25 9.00 9.00 8.00 7.50 8.00 7.75 Semi- 84 MF4SW Mana Mid washed 84.00 48.00 83.50 7.75 7.75 7.50 7.50 7.75 7.75 Semi- 85 744SW Limu Kosa Mid washed 79.00 45.00 81.75 7.50 7.50 7.50 7.25 7.50 7.25 Semi- 86 MF1SW Mana Mid washed 84.00 48.00 84.00 7.75 7.75 7.75 7.50 7.75 7.75 a High altitude: >1750 m asl, Mid altitude: 1550-1750 m asl and Low altitude: <1550 m asl. b Overall cup preference, acidity, body, aroma, flavour and aftertaste are a part of specialty cup quality attributes. Samples 1-66 are calibration samples and samples 67-86 are prediction samples.

115

Chapter 5

Table 5.2: Summary of NIR model parameters for preliminary and specialty cup quality attributes Coffee quality attributes Number of spectra Selected Wavelength range Data pre-processing data points (cm-1) Cross validation Prediction data set (n = 66) data set (n =20) Preliminary cup quality 184 60 906 8916-6940; 6420-5020 & 4252-4016 First derivatives + Vector normalisation Preliminary total quality 195 60 975 8916-5020 Subtraction of straight line Total specialty cup quality 182 60 1009 8944-8524 & 8028-4420 MSC* Specific specialty cup quality attributes Overall cup preference 179 60 1055 10028-10012; 9268-9252; 8968-5468 & First derivatives + MSC 4712-4040 Acidity 173 60 892 1540-11528; 8944-8524 & 7544-4420 Vector normalisation Body 176 60 714 8944-8524; 8028-7124& 5940-4420 First derivatives + MSC Aroma 192 60 619 8944-8524; 6904-6380 & 5940-4420 First derivatives Flavour 185 60 1136 11540-11528 & 8944-4420 First derivatives + Vector Normalisation Aftertaste 180 60 682 11540-11528; 8944-8524; 8028-7268 & First derivatives + 5940-4420 Vector Normalisation *Multiplicative scatter correction

116

Chapter 6

General discussion, conclusions and future perspectives

117

Chapter 6

6 General discussions

The specialty coffee market is constantly growing mainly due to an increased interest in taste and other quality attributes of high quality coffee from some part of the consumers. This growing demand for high quality coffee triggered coffee producing countries to produce specialty coffees for the top-end markets. In Ethiopia, using ‘total specialty cup scores’, top quality coffees are grouped into Q1 (specialty 1, score: ≥85) and Q2 (specialty 2, score: 80 - < 85). This classification resulted in market segmentation and higher prices on the international markets. In 2015-2016, for instance, Q1 and Q2 coffees were sold at ca. $7.7 and $5.9 per kilogram of green coffee beans, respectively. In contrast, grade 3 coffee beans were sold at price of less than ca. $3.6 (40 – 50% less than prices of Q1 and Q2 coffee beans). This indicates that improving coffee quality is a key opportunity for coffee producing countries to attract foreign currency. The premium price of specialty coffee also benefits farmers as the return is given to farmers. Thus, increase in coffee price at international market will lead to increases in return to farmers. However, these varies among farmers e.g. farmers who are a member of cooperatives benefited than the non- member. For instance, an increase in coffee price at international market could lead an increase in price at local market at which the private traders are forced to increase coffee cherry price at local market to compete with cooperatives. As the result non- member of cooperatives will be benefited indirectly.

Coffee quality is, however, determined by numerous factors along its value chain. Genetic traits, growing environment, harvest period, cherry sorting, postharvest processing methods and types of actors involved in the supply chain are known to predominantly affect coffee quality. Poor control over these factors is the major causes of coffee quality loss in the Ethiopian coffee value chain. This requires in depth investigation and analysis of these variables and their possible interactions on chemical compounds (i.e. quality precursors) and final coffee beverage quality. This PhD thesis, therefore, examined key factors: growing environment, harvest period, cherry sorting, postharvest processing methods and types of actors that control coffee bean quality along the coffee value chain and introduced a fast, reliable and objective coffee quality assessment tool. The ultimate outcome was to deliver concepts and management methods that could enhance coffee quality and propose a physical-based method for coffee quality assessment. 118

General discussion, conclusions and future perspectives

The thesis is divided into three different sections: The first two experimental chapters (Chapter 2 and 3) describe the influence of growing environment (altitude and shade), management practices (harvest period and postharvest processing methods) on green bean quality. Chapter 4 describes how different actors in the coffee value chain, coffee cherry sorting and postharvest processing affect coffee bean quality. Chapter 5 describes the development and validation of NIRS- based approach that allows predicting coffee cup quality from NIR spectra of ground green beans. In this final chapter (Chapter 6) the main findings, conclusions and suggestions for future research directions are presented.

6.1 Optimization of biophysical factors to improve coffee bean quality

Growing environment affects quality of green coffee beans. This study investigated effects of altitude and shade combined with harvest period and postharvest processing method on quality and biochemical composition of coffee beans in Mana district, Jimma zone, Ethiopia. Coffee quality increased with increasing altitude (Chapter 2 and 3) and shading (Chapter 2). Coffee beans harvested from ‘high’ altitude i.e.1950 – 2100 m asl. received mostly (ca. 87%) a Q1 grading score. In contrast, all coffee beans harvested from ‘mid’ (1600 – 1650 m asl) and ‘low’ (1450 - 1550 m asl.) altitude were classified as Q2 (Chapter 3). The increase in coffee bean quality with altitude was mainly due to its control on growing temperature; high altitude temperature ranges provide optimal time for bean maturation. Longer bean maturation increases accumulation of flavour precursor compounds such as sugars, chlorogenic acid and is, hence, positively related to coffee bean quality.

The degree of shading has a similar effect. Though the optimal shading depends on altitude. At high altitudes, dense shade (65-85%), however, reduced the share of Q1 coffee beans by ca. 50% while the largest percentage of Q1 coffee (ca. 67%) were obtained from medium shading, followed by open sun condition (ca. 53%). On the other hand, at mid altitude, medium shading (40-55%) allowed to increase the percentage of Q1 slightly from 7 to 13% (Chapter 2). This increase was below expectations and was not significantly different from the percentage of Q1 beans obtained under open sun and dense shading. The reason for such phenomena was not very clear but it might be explained by leaf fall of shade trees (i.e. deciduous nature of shade tree types) observed

119

Chapter 6

particularly during the periods of coffee cherry maturation. This indicates that farmers should consider the best adapted native shade trees to grow coffee under consistent shade level and to increase Q1 coffee bean. In general, this approach could also help to adapt to an increasing temperatures due to climate change.

Both altitude and shade also significantly influenced total chlorogenic acid content of green coffee beans. Mid altitude and open sun growing conditions favour the accumulation of total chlorogenic acid (46.5 g kg -1). Coffee beans grown at high altitude and open sun, on the other hand, had the lowest total chlorogenic acid content (40.5 g kg -1). Caffeine content, on the other hand, was significantly affected by altitude, shade and harvest period interactions. Early harvested beans from mid altitude grown and dense shade contained the highest amount of caffeine (17.9 g kg -1), while coffee beans grown at high altitude and medium shade but harvested at mid season harvest had the lowest caffeine content (14.5 g kg -1). However, in chapter 3 caffeine, total chlorogenic acid and sucrose content increased with altitude by 11, 9, and 10%, respectively. The reason of this discrepancy could be due to differences in year or seasons and factors (e.g. shade, harvest period and postharvest processing methods) considered in both studies.

Regardless of growing altitude, this study also showed that harvesting periods (Chapter 2) and coffee processing methods (Chapter 3 and 4) affect coffee bean quality. Early to middle harvest period improved coffee bean quality. Compared to late harvesting, early harvesting increased the percentage Q1 coffee from 27 to 73%, at high, and from 7 to 20% at mid altitude. Similarly, dry processing method improved coffee bean quality and increased the percentage of coffee beans falling into Q1 grades by more than 50% compared to washed processing method (Chapter 4). The existing perception in Ethiopia is that the top specialty coffees should be produced via washed rather than dry processing. However, quality differences between washed and dry processed beans should be rather explained by the intrinsic quality of coffee cherries considered for processing. In actual practice, only red ripe and clean coffee cherries are used for washed processing method while a mixture of immature, ripe and overripe cherries are usually processed via dry processing method. This leads to substantial quality losses in dry processed coffee beans. In contrast, for selectively harvested and properly sorted coffee beans, our result showed that dry processing method results in beans with higher or equal coffee beverage quality. For instance, for sorted coffee cherries, all dry

120

General discussion, conclusions and future perspectives

processed coffee beans samples qualified for Q1 while only ca. 33% of washed processed coffee beans fall into Q1 (Chapter 4). In Ethiopia the current share of coffee beans produced by washed method is about 25%, while the remaining 75% are dry processed coffee beans. This might therefore limit Ethiopia’s high quality coffee export share in the international specialty market. Furthermore, caffeine and sucrose were significantly affected by altitude and processing interactions. High altitude grown coffee beans processed via any of processing methods gave higher caffeine content than low altitude grown coffee beans. Moreover, at high altitude, washed processed coffee beans gave lower sucrose content compared to dry and semi-washed beans processed coffee beans. At mid and low altitude, on the other hand, processing methods did not show significant differences on any biochemical compounds (Chapter 3). In general this study indicated altitude was the main driving factor that affects biochemical compositions of green coffee beans.

Furthermore, the effect of altitude, shade and postharvest processing on coffee quality and biochemical compositions may also varies with genotypes. Des Marais et al. (2013) reported the existence interaction between genotype and environment in many plants. Thus, positive interaction of genotype and a particular environment results in better quality coffee. Bote (2016), on the other hand, reported that growing environment e.g. altitude had a stronger effect on Ethiopian coffee quality compared the genotypes (Bote, 2016). Still these small differences could contribute to bring change in coffee quality grade (Q1 vs. Q2) and future breeding research should consider genotypes by environment for quality improvement.

6.2 Coffee quality along the value chain: evidence from farmers, cooperatives and private traders

The predominant coffee actors that influence coffee quality in the supply chain are smallholder coffee farmers, cooperatives and private traders. Farmers produce coffee beans while cooperatives and private traders process and export coffee beans. This thesis examined impacts of cooperatives, private traders and coffee certification on the quality of green coffee beans in Jimma zone, southwestern Ethiopia (Chapter 4). The result of this research generally showed that coffees passing through cooperatives had superior bean qualities compared to coffees passing through private traders. About 33 121

Chapter 6

and 67% of cooperative coffee qualified as Q1 and Q2, respectively. Coffee beans collected from private traders classified mainly into commercial grade 3 (ca. 78%), only 22% of the samples were qualified as Q2 and none as Q1. Quality differences between cooperative and private trader coffee beans could be explained by the differences in type of coffee cherries (red ripe, unripe and overripe) delivered by farmers and growing environment. In coffee cooperatives, coffee cherries harvested from different production sites are less likely mixed with each other and are processed separately, while private traders collect coffee cherries from different growing environments (i.e. altitude) and process bulk coffee beans regardless of coffee production sites. Majority of coffee beans processed by cooperatives qualified for specialty coffees which was not the case for private trader. This shows that a site-by-site processing approach (processing coffee harvested from similar growing environments) contributes to improved bean quality and links bean characteristics with coffee bean production sites. But, there is still more room (e.g. via sorting) for cooperatives to improve the percentage of Q1 coffee beans over Q2. Sorting coffee cherries before processing increased the percentage Q1 coffee beans from 17 to 61%. Sorting along with site-by-site coffee processing approach would help actors to supply a consistently high quality product. Moreover, initiating further establishment of cooperatives would also allow generating better traceable coffee beans.

Apart from quality improvement, the establishment of cooperatives in the study area also enhanced coffee price paid by private traders. In an area without cooperatives (e.g. in Buxure Peasant Association in Mana district, Jimma zone) farmers receive low prices ($ 0.3 per kilogram of red cherries) compared to areas with cooperatives ($ 0.5 per kilogram of red cherries). This, therefore, calls for the involvement of institutions that control local market price and deliver price information to farmers. An unexpected result from this study was that coffee certification did not result in any quality improvement. This is probably due to the fact that certification mainly focuses on promotion of socioeconomic advantage and environmental sustainability rather than coffee quality. It therefore requires more effort to promote certification along with quality improvement as the demand for certified coffee is increasing.

122

General discussion, conclusions and future perspectives

6.3 Need for improved coffee quality assessment method

Nowadays, the growing global demand for specialty coffee warrants the need for improved coffee quality assessment methods. In Ethiopia, and elsewhere, green bean coffee quality is judged base on physical (e.g. black beans, immature beans) and cup quality (e.g. acidity and flavor) attributes. Every coffee growing region in Ethiopia has several coffee tasting experts, so-called 'coffee cuppers'. Coffee cuppers are trained experts that adhere to strict guidelines to ensure coffee quality for export and local market. These evaluation methods are, however, subjective, costly, time consuming, and require meticulous sample preparation. This calls for the development of a rapid, low- cost, reliable and reproducible analytical method to evaluate coffee quality attributes. Near infrared spectroscopy (NIRS) is one of the increasingly growing techniques in food analysis. It has also been applied to detect coffee adulteration, discriminate arabica from Robusta, detect defective beans, determine chemical composition such as caffeine content, predict roasting degree, and even cup quality attributes such as acidity and body.

In this PhD research, a NIR-based model using green bean NIR spectra was developed to predict cup quality attributes of Ethiopian arabica coffee beans (Chapter 5). The developed model had a very good predictive power for total specialty cup quality and most of its specific quality attributes. For instance, the correlation coefficients (r) between measured and predicted cupping scores for total specialty cup quality were 90. The developed model is, therefore, a promising tool for fast and accurate assessment of specialty coffee cup quality. Coffee industry is growing in Ethiopia and huge numbers of coffee samples are daily delivered to the ECX laboratories for quality analysis and grading. Our NIR- based approach provides could greatly contributes to introduce a fast and reliable coffee quality assessment and classification of the samples into Q1 (≥ 85), Q2 (80 – < 85) and commercial grade 3 (71 - < 80). The method indeed simplifies coffee grading systems in the country and reduces the time needed to grade beans by the conventional method. More recently, international buyers or consumers, are also increasingly demanding transparency and accountability along the supply chains. Thus, our model could also contributes to ensure quality of sold coffee beans that increases consumer confidence in the quality of the product. This, however, requires institutional (e.g. ECX) decision to integrate this new tool into the current cup tasting system. In

123

Chapter 6

addition, the NIRS approach can also be used to asses coffee quality at cooperative or private trader’s level. These enable them to quickly inform the farmers on the quality of the coffee they produce and to identify where exactly quality is lost in the value chain. This also encourages farmers to produce and deliver good quality coffee cherries to their respective cooperatives or private trader. Overall, the method would also enhance and accelerates price paid to the farmers.

6.4 Final conclusions

This PhD research, in general, indicated that coffee beans from high altitude grown in open or medium shade and harvested early to mid season gave higher quality. Dense shade (65-85%) at high altitude, however, causes dramatic reduction in the share of Q1 coffee by 50%. This shows that coffee producers at high alitiude should avoid dense shade e.g.via pruning of the shade trees. Thus, to maintain the quality of high altitude grown coffee, producers or processor should avoid mixing of with low alitude coffee.

At mid altitude, on the other hand, no signifcant effect of shading was observed on coffee quality, however, medium shading condition showed a minor increase in the share of Q1 coffee compare to open and dense shading. In general, growing coffee trees at higher altitude or under tailored shading conditions (i.e. for mid to low altitude) could, therefore, be used as a possible approach to adapt to effects of high temperature and to sustain production of high quality coffee beans.

In addition to altitude and shade trees, early to mid season harvest and dry processing improved coffee bean quality. At both high and low altitudes, large percentage of coffee beans harvested early to mid season harvest qualified as Q1 coffees. Similarly, dry processing positively influenced coffee quality and increased the percentage of Q1 coffee beans by 50% for unsorted coffee cherries than washed coffee. Furthermore, sorting coffee cherries before dry processing enhanced further the share of Q1 coffee beans. Thus, almost all sorted coffee beans fell into Q1 coffee. Compared to washed processing method, dry processing method is cheaper and is a natural way of coffee processing method. It also preserves other ecosystem services and prevents environmental pollutions. It is the best processing approach that should be used in the study area. Further improvement of quality of dry processed coffee beans, however, 124

General discussion, conclusions and future perspectives

requires selective picking and sorting of coffee cherries prior to subjecting to dry processing method. In addition, appropriate drying material (e.g. mesh wire and place (raised bed) should need due attention to maintain coffee quality during drying. It also needs institutional interference to incentives farmers for high quality produces and create awareness about coffee quality of dry processing method.

Biochemical compostion ( e.g. total chlorgenic acid) of green coffee beans was also found to be influenced by altitude and shading interaction while caffeine content was affected by interaction between altitude, shade and havest period. Both caffeine and total chlorgenic acid showed slight reduction with altitude (Chapter 2). Coffee beans grown at high altitude and processed via dry and semi-washed methods gave higher sucrose contents compared to washed coffee beans (Chapter 3). However, processing methods did not show significant differences on any of biochemical compounds at mid and low altitude grown coffee beans. The caffeine, total chlorogenic acid and sucrose contents in green coffee beans, on the other hand, increased with altitude (Chapter 3). In general, the result of both study (Chapter 2 and 3) were not consistent for these chemical contents and requires further studies to give clear conclusions.

The establishment of cooperatives in the coffee value chain improves coffee bean quality and farmers’ livelihood via improving coffee prices and information transfer. Cooperatives processes coffee beans site-by-site and do not bulk coffee cherries from different production sites i.e. different growing environments. This approach helps consumers and all actors in the coffee chain to link the bean quality characteristics with production site or origin. Apart from improving bean quality, enhancing organization of coffee smallholders into cooperatives could improve Ethiopian’s export shares in the international market and livelihoods of smallholders via better prices paid to farmers. Proper harvesting and sorting of coffee cherries help cooperatives and private traders to further maintain coffee quality and supply consistently high quality products to the top-end markets.

Overall, all biophysical factors and actors in the coffee value chain affect coffee bean quality. Small changes of quality scores driven by these factors along the value chain induce dramatic switches in the fraction Q1 vs. Q2 coffee beans and affects smallholders’ livelihood and competitiveness in the international markets. The study also developed and proposed NIRS-based approach for coffee quality assessment. The

125

Chapter 6

approach could provide a fast, easy and accurate coffee quality assessment techniques which can be used at ECX to assist the current coffee quality testing and grading system. All together the study identified important biophysical factors (e.g. altitude, shade), management practices (e.g. harvest period, sorting and postharvest processing) and actors in the value chain that control coffee bean quality, opened concepts and methodologies for enhancing coffee bean quality and developed a promising tool for fast and accurate assessment of specialty coffee cup quality.

6.5 Practical implications

This thesis identified key biophysical factors (e.g. altitude, shade), management practices (e.g. harvest period, sorting and postharvest processing) and actors in the value chain that control coffee bean quality. Here we attempt to translate our findings into practical management considerations.

Growing of coffee tree at high altitude with open to medium shade and harvested at early to middle season suggested as good management techniques to produce top specialty coffee and to increase share of Ethiopian coffee on specialty of market. However, at high altitude, dense shade can cause dramatic quality losses and producers of this area should reduce the shade level via pruning. At both altitudes early and middle harvesting can be suggested to produce more Q1 coffee. Thus, to maintain high quality of high altitude and early to middle harvest beans needed to be kept separately from time of harvest to marketing.

Moreover, dry processing positively influences coffee quality and allow increasing Q1 coffee. In addition, because of easy management, low production cost and sustainability, dry processing should be recommended over washing methods. Regardless of growing conditions, use of red ripe coffee cherries via sorting or proper harvest significantly improved coffee quality, hence, special care should be given during coffee harvesting or sorting is recommend after harvest. The establishment of cooperatives in the coffee value chain improves coffee bean quality. So governmental institution should encourage and enhance organization of coffee smallholders into cooperatives. In general all contributing factors along value chain should be controlled to improve and maintain coffee quality. Finally the developed NIR- based model is a promising tool for fast and

126

General discussion, conclusions and future perspectives

accurate assessment (i.e. Q1 vs. Q2) of specialty coffee quality and to support the current coffee quality testing and grading system.

6.6 Perspectives for future research

Although this thesis identified important biophysical factors (e.g. altitude, shade) and introduced advanced quality assessment techniques, the results also indicated some prospective for further research to improve Ethiopian coffee quality. In Chapter 2, influence of altitude, shade and harvest period on coffee bean quality was investigated and optimum shade level and harvest period were identified for each altitude. The study, however, did not investigate year-by-year variations and soil type effects. Future studies need to integrate these factors and analyse their impacts on the beverage characteristics of coffee beans. Moreover, in southwest Ethiopia e.g. Bebeka, Arabica coffee trees are also grown at very low altitudes < 1400 m asl. Future studies should also consider such production areas to identify optimum shade levels and optimum harvest periods. Further studies also needed to be initiated to identify best adapted shade tree types (i.e. trees that do not shade their leaves) to produce coffees with consistently high bean quality at mid and at low altitude areas. In addition, more attention should be given to the endemic or native adapted shade tree types because, use of the exotic shade tree types may limit coffee certification scheme (e.g. Rainforest alliance).

The performance of coffee trees across different altitudinal gradient and shade levels varies from variety to variety. This implies that the same variety grown in different environment may vary significantly in quality. Thus, variety grown in place other than their preferred growing environments, may not give coffees with good taste and flavour. This requires future studies that evaluate variety by environment (altitude and shade) interaction on Ethiopian coffee bean quality and identify ideal growing environment. Moreover, exploring or identifying genes that are responsible for superior quality coffee and examining the response of these genes to different growing environment are also the ongoing questions.

Although caffeine, chlorogenic acid and sucrose content of green coffee beans in relation to growing environment (altitude and shade), harvest periods and postharvest processing methods were examined in this study, further research should also be initiated to 127

Chapter 6

evaluate effects of these factors on lipid, protein, trigonelline and other important compounds found in coffee beans.

In chapter 4 variability of coffee quality among different coffee actors along the value chain was studied and the results gave better understandings on how coffee actors (farmers, cooperatives and private traders) manage coffee beans. To further map Ethiopian coffee quality along value chain i.e. from farmer to exporter, other similar studies including private companies and unions should be initiated and studied in other coffee growing parts of the country. NIRS- based model was developed to assess coffee cup quality in this study. The model, however, need to be further tested using large data sets and for coffee samples collected from different regions in Ethiopia. This would allow examining if one generic or a region-specific model is needed across the country. Exploring NIRS application for coffee origin discrimination, traceability and biochemical analysis are also areas to be addressed in the future.

128

References

Abasanbi, A. (2010). Assessment of coffee quality and its related problems in Jimma Zone of Oromia Regional State. Abeyot, T., Sentayehu , A., Taye, K., & Weyessa, G. (2011). Genetic diversity analysis for quality attributes of some promising Coffea arabica germplasm collections in southwestern Ethiopia. Journal of Biological Sciences, 11(3), 236-244. Abreu, H. M. C. d., Nobile, P. M., Shimizu, M. M., Yamamoto, P. Y., Silva, E. A., Colombo, C. A., & Mazzafera, P. (2012). Influence of air temperature on proteinase activity and beverage quality in Coffea arabica. Brazilian Journal of Botany, 35(4), 357-376. Adam, K. (2009). Sustainable coffee certifications a comparison matrix (pp. 1 - 5). Aerts, R., Geeraert, L., Berecha, G., Hundera, K., Muys, B., De Kort, H., & Honnay, O. (2017). Conserving wild Arabica coffee: Emerging threats and opportunities. Agriculture, Ecosystems & Environment, 237, 75-79. Aerts, R., Hundera, K., Berecha, G., Gijbels, P., Baeten, M., Van Mechelen, M., . . . Honnay, O. (2011). Semi-forest coffee cultivation and the conservation of Ethiopian Afromontane rainforest fragments. Forest Ecology and Management, 261(6), 1034-1041. Aerts, R. J., & Baumann, T. W. (1994). Distribution and utilization of chlorogenic acid in Coffea seedlings. Journal of Experimental Botany, 45(273), 497-503. Aeschbacher, H. U., Wolleb, U., Löliger, J., Spadone, J. C., & Liardon, R. (1989). Contribution of coffee aroma constituents to the mutagenicity of coffee. Food and Chemical Toxicology, 27(4), 227-232. Agwanda, C., Baradat, P., Eskes, A., Cilas, C., & Charrier, A. (2003). Selection for bean and liquor qualities within related hybrids of Arabica coffee in multilocal field trials. Euphytica, 131(1), 1-14. Agwanda, C. O., Lashermes, P., Trouslot, P., Combes, M.-C., & Charrier, A. (1997). Identification of RAPD markers for resistance to coffee berry disease, Colletotrichum kahawae, in arabica coffee. Euphytica, 97(2), 241-248. Alemu, M. M. (2015). Effect of Tree Shade on Coffee Crop Production. Journal of Sustainable Development, 8(9), 66. Alonso-Salces, Maria, R., Guillou, C., & Berrueta, L. A. (2009). Liquid chromatography coupled with ultraviolet absorbance detection, electrospray ionization, collision induced dissociation and tandem mass spectrometry on a triple quadrupole for the on line characterization of polyphenols and methylxanthines in green coffee beans. Rapid Communications in Mass Spectrometry, 23(3), 363-383. Alonso-Salces, Rosa M, Serra, F., Reniero, F., & HÉberger, K. (2009). Botanical and geographical characterization of green coffee (Coffea arabica and Coffea canephora): chemometric evaluation of phenolic and methylxanthine contents. Journal of agricultural and food chemistry, 57(10), 4224-4235. Amézqueta, S., González-Peñas, E., Murillo-Arbizu, M., & de Cerain, A. L. (2009). Ochratoxin A decontamination: A review. Food Control, 20(4), 326-333. Anese, M., Quarta, B., & Frias, J. (2011). Modelling the effect of asparaginase in reducing acrylamide formation in biscuits. Food chemistry, 126(2), 435-440. Anthony, F., Bertrand, B., Quiros, O., Wilches, A., Lashermes, P., Berthaud, J., & Charrier, A. (2001). Genetic diversity of wild coffee (Coffea arabica L.) using molecular markers. Euphytica, 118(1), 53-65.

129

References

Anthony, F., Combes, M. C., Astorga, C., Bertrand, B., Graziosi, G., & Lashermes, P. (2002). The origin of cultivated Coffea arabica L. varieties revealed by AFLP and SSR markers. Theoretical and Applied Genetics, 104(5), 894-900. AOAC. (2000). Official methods of analysis of the association of official analytical chemists, 17th (eds.). Association of Official Analytical,Gaithersburg, MD, USA. ARDO. (2008). Annual Report of Agriculture and Rural Development Office of Manna woreda,Yebbu, Manna. Addis Ababa, Ethiopia. Arion, W. J., Canfield, W. K., Ramos, F. C., Schindler, P. W., Burger, H.-J., Hemmerle, H., Herling, A. W. (1997). Chlorogenic acid and hydroxynitrobenzaldehyde: new inhibitors of hepatic glucose 6-phosphatase. Archives of biochemistry and biophysics, 339(2), 315-322. Ashihara, H., & Crozier, A. (2001). Caffeine: a well known but little mentioned compound in plant science. Trends in plant science, 6(9), 407-413. Ashihara, H., Sano, H., & Crozier, A. (2008). Caffeine and related purine alkaloids: biosynthesis, catabolism, function and genetic engineering. Phytochemistry, 69(4), 841-856. Avallone, S., Guyot, B., Brillouet, J.-M., Olguin, E., & Guiraud, J.-P. (2001). Microbiological and biochemical study of coffee fermentation. Current microbiology, 42(4), 252-256. Avelino, J., Barboza, B., Araya, J. C., Fonseca, C., Davrieux, F., Guyot, B., & Cilas, C. (2005). Effects of slope exposure, altitude and yield on coffee quality in two altitude terroirs of Costa Rica, Orosi and Santa Maria de Dota. Journal of the Science of Food and Agriculture, 85(11), 1869-1876. Avelino, J., Barboza, B., Davrieux, F., & Guyot, B. (2007). Shade effect on sensory and chemical characteristics of coffee from very high altitude plantations in Costa Rica. In:Second International Symposium on Multi-strata Agroforestry Systems with Perennial Crops, 17–21 September 2007, IUFRO, CIRAD, CATIE and ICRAF,Turrialba, Costa Rica. Bacon, C. (2005). Confronting the coffee crisis: can fair trade, organic, and specialty coffees reduce small-scale farmer vulnerability in northern Nicaragua? World development, 33(3), 497-511. Baldwin, E. A., Bai, J., Plotto, A., & Dea, S. (2011). Electronic noses and tongues: Applications for the food and pharmaceutical industries. Sensors, 11(5), 4744- 4766. Balyaya, K., & Clifford, M. N. (1995). Chlorogenic acids and caffeine contents of monsooned Indian Arabica and Robusta coffees compared with wet and dry processed coffees from the same geographic area. In: ASIC, 16 e Colloque, Kyoto. Baptista, P., Felizardo, P., Menezes, J. C., & Correia, M. J. N. (2008). Multivariate near infrared spectroscopy models for predicting the iodine value, CFPP, kinematic viscosity at 40 C and density at 15 C of biodiesel. Talanta, 77(1), 144-151. .Barbin, D. F., Felicio, A. L. d. S. M., Sun, D.-W., Nixdorf, S. L., & Hirooka, E. Y. (2014). Application of infrared spectral techniques on quality and compositional attributes of coffee: An overview. Food Research International, 61(0), 23-32. Barbosa, J., Borem, F., Alves, H., Cirillo, M., Sartori, M., & Ducatti, C. (2014). Discrimination of Production Environments of Specialty Coffees by Means of Stable Isotopes and Discriminant Model. Journal of Agricultural Science, 6(5), p55. Bart, M., Seneshaw, T., Tadesse, K., & Yaw, N. (2014). Structure and performance of Ethiopia’s coffee export sector. ESSP II Working Paper 66. Washington, D.C. and Addis Ababa, Ethiopia: International Food Policy Research Institute (IFPRI) and Ethiopian Development Research Institute (EDRI).

130 References

Bayetta. (2001). Arabica coffee breeding for yield and resistance to coffee berry disease (colletotrichum kahawae sp. nov.) Beer, J., Muschler, R., Kass, D., & Somarriba, E. (1998). Shade management in coffee and cacao plantations. In Directions in Tropical Agroforestry Research, (pp. 139- 164): Springer. Bekele, Y. D. (2005). Assessment of cup quality, morphological, biochemical and molecular diversity of Coffea arabica L. genotypes of Ethiopia. University of the Free State. Belete, Y., Belachew, B., & Fininsa, C. (2014). Evaluation of bean qualities of indigenous Arabica coffee genotypes across different environments. Journal of Plant Breeding and Crop Science, 6(10), 135-143. Belitz, H., Grosch, W., & Schieberle, P. (2009). Food Chemistry, forth revised and extended edition. In): Berlin, Heidelberg: Springer-Verlag. Belkaid, A., Currie, J.-C., Desgagnés, J., & Annabi, B. (2006). The chemopreventive properties of chlorogenic acid reveal a potential new role for the microsomal glucose-6-phosphate translocase in brain tumor progression. Cancer Cell International, 6(1), 7. Bello-Mendoza, R., & Castillo-Rivera, M. (1998). Start-up of an anaerobic hybrid (UASB/Filter) reactor treating wastewater from a coffee processing plant. Anaerobe, 4(5), 219-225. Berecha, G., Aerts, R., Vandepitte, K., Van Glabeke, S., Muys, B., Roldán-Ruiz, I., & Honnay, O. (2014). Effects of forest management on mating patterns, pollen flow and intergenerational transfer of genetic diversity in wild Arabica coffee (Coffea arabica L.) from Afromontane rainforests. Biological Journal of the Linnean Society, 112(1), 76-88. Bernard, T., & Spielman, D. J. (2009). Reaching the rural poor through rural producer organizations? A study of agricultural marketing cooperatives in Ethiopia. Food Policy, 34(1), 60-69. Berthaud, J., & Charrier, A. (1988). Genetic resources of Coffea. Bertrand, B., Boulanger, R., Dussert, S., Laffargue, A., Ribeyre, F., Berthiot, L., Joët, T. (2012). Climatic factors directly impact the biochemical composition and the volatile organic compounds fingerprint in green Arabica coffee bean as well coffee beverage quality. In International Conference on Coffee Science, (pp. 628--635). Bertrand, B., Etienne, H., Guyot, B., & Vaast, P. (2005). Year of production and canopy region influence bean characteristics and beverage quality of Arabica coffee. In: ASIC, 20th International Conference on Coffee Science, Bangalore, India,11-15 October 2004., (pp.878-886). Bertrand, B., Vaast, P., Alpizar, E., Etienne, H., Davrieux, F., & Charmetant, P. (2006). Comparison of bean biochemical composition and beverage quality of Arabica hybrids involving Sudanese-Ethiopian origins with traditional varieties at various elevations in Central America. Tree physiology, 26(9), 1239-1248. Bleibaum, R. N., Stone, H., Tan, T., Labreche, S., Saint-Martin, E., & Isz, S. (2002). Comparison of sensory and consumer results with electronic nose and tongue sensors for apple juices. Food Quality and Preference, 13(6), 409-422. Boisseau, M., Aribi, J., Sousa, F. R. d., Carneiro, R. M., & Anthony, F. (2009). Resistance to Meloidogyne paranaensis in wild Coffea arabica. Tropical Plant Pathology, 34(1), 38-41. Borém, F. M., Ribeiro, D. R., Taveira, J. H. S., Prado, M. V. B., Ferraz, V., Tosta, M. F Luz ,M.P.S. Cortez, R.M. Shuler, J. (2014). Genotype and Environment Interaction in Chemical Composition and Sensory Quality of Natural Coffees. In: Proceeding of the 25th international congress of ASIC.

131 References

Bosselmann, A., Dons, K., Oberthur, T., Olsen, C. S., Ræbild, A., & Usma, H. (2009). The influence of shade trees on coffee quality in small holder coffee agroforestry systems in Southern Colombia. Agriculture, Ecosystems & Environment, 129(1), 253-260. Bote, & Struik, P. C. (2011). Effects of shade on growth, production and quality of coffee (Coffea arabica) in Ethiopia. Journal of Horticulture and Forestry, 3(11), 336 - 341. Bote, A. (2016). Examining growth, yield and bean quality of Ethiopian coffee trees. Wageningen University. Bote, A. D., & Vos, J. (2016). Erratum to: Branch growth dynamics, photosynthesis, yield and bean size distribution in response to fruit load manipulation in coffee trees. Trees, 30(5), 1911-1911. Bressanello, D., Liberto, E., Cordero, C., Rubiolo, P., Pellegrino, G., Ruosi, M. R., & Bicchi, C. (2017). Coffee aroma: Chemometric comparison of the chemical information provided by three different samplings combined with GC–MS to describe the sensory properties in cup. Food chemistry, 214, 218-226. Briandet, R., Kemsley, E. K., & Wilson, R. H. (1996). Approaches to Adulteration Detection in Instant Coffees using Infrared Spectroscopy and Chemometrics. Journal of the Science of Food and Agriculture, 71(3), 359-366. Buffo, R. A., & Cardelli -Freire, C. (2004). Coffee ﬂavour: an overview. Flavour and fragrance journal, 19(2), 99-104. Bunn, C., Läderach, P., Rivera, O. O., & Kirschke, D. (2015). A bitter cup: climate change profile of global production of Arabica and Robusta coffee. Climatic Change, 129(1-2), 89-101. Buratti, S., Sinelli, N., Bertone, E., Venturello, A., Casiraghi, E., & Geobaldo, F. (2014). Discrimination between washed Arabica, natural Arabica and Robusta coffees by using near infrared spectroscopy, electronic nose and electronic tongue analysis. Journal of the Science of Food and Agriculture. Buratti, S., Sinelli, N., Bertone, E., Venturello, A., Casiraghi, E., & Geobaldo, F. (2015). Discrimination between washed Arabica, natural Arabica and Robusta coffees by using near infrared spectroscopy, electronic nose and electronic tongue analysis. Journal of the Science of Food and Agriculture, 95(11), 2192-2200. Byers, A., Giovannucci, D., & Liu, P. (2008). Value-adding standards in the North American food market. Trade opportunities in certified products for developing countries: FAO. Bytof, G., Knopp, S., Kramer, D., Breitenstein, B., Bergervoet, J., Groot, S. P., & Selmar, D. (2007). Transient occurrence of seed germination processes during coffee post-harvest treatment. Annals of Botany, 100(1), 61-66. Bytof, G., Knopp, S., Schieberle, P., Teutsch, I., & Selmar, D. (2005). Influence of processing on the generation of γ-aminobutyric acid in green coffee beans. European Food Research and Technology, 220(3-4), 245-250. Camargo, A., Santinato, R., & Cortez, J. (1992). Aptidao climática para qualidade da bebida nas principais regioes cafeeiras de café Arábica. In: Congresso Brasileiro de Pesquisas Cafeeiras, 18. Araxá (Brasil), Outubro 27-30, 1992. Trabalhos apresentados. Cannell, M. (1985). Physiology of the coffee crop. In: Coffee (pp. 108-134): Springer. Carr, M.K.V. (2001). The water relations and irrigation requirements of coffee. Experimental Agriculture, 37(1), 1-36. Cheng, B., Furtado, A., Smyth, H. E., & Henry, R. J. (2016). Influence of genotype and environment on coffee quality. Trends in Food Science & Technology.

132 References

Cho, A.-S., Jeon, S.-M., Kim, M.-J., Yeo, J., Seo, K.-I., Choi, M.-S., & Lee, M.-K. (2010). Chlorogenic acid exhibits anti-obesity property and improves lipid metabolism in high-fat diet-induced-obese mice. Food and Chemical Toxicology, 48(3), 937-943. Clifford. (1999). Chlorogenic acids and other cinnamates–nature, occurrence and dietary burden. Journal of the Science of Food and Agriculture, 79, 362-372. Clifford. (2000). Chlorogenic acids and other cinnamates–nature, occurrence, dietary burden, absorption and metabolism. Journal of the Science of Food and Agriculture, 80(7), 1033-1043. Clifford, Johnston, K. L., Knight, S., & Kuhnert, N. (2003). Hierarchical scheme for LC- MS n identification of chlorogenic acids. Journal of agricultural and food chemistry, 51(10), 2900-2911. Clifford, & Kazi, T. (1987). The influence of coffee bean maturity on the content of chlorogenic acids, caffeine and trigonelline. Food chemistry, 26(1), 59-69. Clifford, & Wilson, K. (1985). Coffee, Botany. Biochemistry and Production of Beans and Beverage, 305. Consuelo, P., Isabel, E.-D., Josea-Maria, G. z.-s., & Michele, F. (2007). Use of Near- Infrared Spectroscopy and Feature Selection Techniques for Predicting the Caffeine Content and Roasting Color in Roasted Coffees. J. Agric. Food Chem, 55, 7477-7488. Cortez, J. (1997). Aptidão climática para a qualidade da bebida nas principais regiões cafeeiras de Minas Gerais. Informe Agropecuário, 18(187), 27-31. Coste, R., Cambrony, H., & Tindall, H. (1992). Coffee: the plant and the product: Macmillan London. Craig, A. P., Franca, A. S., & Oliveira, L. S. (2011). Discrimination between Immature and Mature Green Coffees by Attenuated Total Reflectance and Diffuse Reflectance Fourier Transform Infrared Spectroscopy. Journal of Food Science, 76(8), C1162-C1168. Craig, A. P., Franca, A. S., & Oliveira, L. S. (2012). Discrimination between defective and non-defective roasted coffees by diffuse reflectance infrared Fourier transform spectroscopy. LWT-Food Science and Technology, 47(2), 505-511. Craig, A. P., Franca, A. S., & Oliveira, L. S. (2012). Evaluation of the potential of FTIR and chemometrics for separation between defective and non-defective coffees. Food chemistry, 132(3), 1368-1374. Craig, A. P., Franca, A. S., Oliveira, L. S., Irudayaraj, J., & Ileleji, K. (2015). Fourier transform infrared spectroscopy and near infrared spectroscopy for the quantification of defects in roasted coffees. Talanta, 134, 379-386. Cubry, P., Musoli, P., Legnate, H., Pot, D., De Bellis, F., Poncet, V.,Leroy, T. (2007). Diversity in coffee assessed with SSR markers: structure of the genus Coffea and perspectives for breeding. Genome, 51(1), 50-63. Curzi, D., & Pacca, L. (2015). Price, quality and trade costs in the food sector. Food Policy, 55, 147-158. Curzi, D., Raimondi, V., & Olper, A. (2014). Quality upgrading, competition and trade policy: evidence from the agri-food sector. European Review of Agricultural Economics. DaMatta, F. M. (2004). Ecophysiological constraints on the production of shaded and unshaded coffee: A review. Field Crops Research, 86(2-3), 99-114. Daniele, G., Alice Byers, Pascal Liu. (2008). Adding Value: Certified Coffee Trade in North America. In Pascal Liu (Ed.) Value-adding Standards in the North American Food Market - Trade Opportunities in Certifi ed Products for Developing Countries. 33-49.

133 References

Davis, A. P., Gole, T. W., Baena, S., & Moat, J. (2012). The impact of climate change on indigenous arabica coffee (Coffea arabica): predicting future trends and identifying priorities. PLoS One, 7(11), e47981. Davis, A. P., Govaerts, R., Bridson, D. M., & Stoffelen, P. (2006). An annotated taxonomic conspectus of the genus Coffea (Rubiaceae). Botanical Journal of the Linnean Society, 152(4), 465-512. Davis, A. P., & Rakotonasolo, F. (2001). Two new species of Coffea L.(Rubiaceae) from northern Madagascar. Adansonia, 23(2), 337-345. De Beenhouwer, M., Geeraert, L., Mertens, J., Van Geel, M., Aerts, R., Vanderhaegen, K., & Honnay, O. (2016). Biodiversity and carbon storage co-benefits of coffee agroforestry across a gradient of increasing management intensity in the SW Ethiopian highlands. Agriculture, Ecosystems & Environment, 222, 193-199. De Beenhouwer, M., Muleta, D., Peeters, B., Van Geel, M., Lievens, B., & Honnay, O. (2015b). DNA pyrosequencing evidence for large diversity differences between natural and managed coffee mycorrhizal fungal communities. Agronomy for sustainable development, 35(1), 241-249. De Castro, R. D., & Marraccini, P. (2006). Cytology, biochemistry and molecular changes during coffee fruit development. Brazilian Journal of Plant Physiology, 18(1), 175- 199. De Maria, C., Trugo, L., Moreira, R., & Petracco, M. (1995). Simultaneous determination of total chlorogenic acid, trigonelline and caffeine in green coffee samples by high performance gel filtration chromatography. Food chemistry, 52(4), 447-449. De Maria, C. A. B., Trugo, L. C., Moreira, R. F. A., & Werneck, C. C. (1994). Composition of green coffee fractions and their contribution to the volatile profile formed during roasting. Food chemistry, 50(2), 141-145. de Melo Pereira, G. V., Soccol, V. T., Pandey, A., Medeiros, A. B. P., Lara, J. M. R. A., Gollo, A. L., & Soccol, C. R. (2014). Isolation, selection and evaluation of yeasts for use in fermentation of coffee beans by the wet process. International journal of food microbiology, 188, 60-66. de Sotillo, D. V. R., Hadley, M., & Sotillo, J. E. (2006). Insulin receptor exon 11+/− is expressed in Zucker (fa/fa) rats, and chlorogenic acid modifies their plasma insulin and liver protein and DNA. The Journal of nutritional biochemistry, 17(1), 63-71. Decazy, F., Avelino, J., Guyot, B., Perriot, J. J., Pineda, C., & Cilas, C. (2003). Quality of Different Honduran Coffees in Relation to Several Environments. Journal of Food Science, 68(7), 2356-2361. Deepa, G., Chanakya, H., de Alwis, A., Manj unath, G., & Devi, V. (2002). Overcoming pollution of lakes and water bodies due to coffee pulping activities with appropriate technology solutions. In: Proceedings “Symposium on Conservation, Restoration and Management of Aquatic Ecosystems”, paper, vol. 4). Dejen, D., & and Matthews, H. (2016). Agricultural Cooperatives, Opportunities and Challenges, the Case of Bench Maji Zone, Ethiopia. Journal of Poverty, Investment and Development, 22. Dempsey, J.(2006). A Case Study of Institution Building & Value Chain Strengthening to Link Ethiopian Cooperative Coffee Producers to International Markets. Addis Ababa, Ethiopia Dempsey, & Campbell, R. (2006). A Value Chain Approach to Coffee Production: Linking Ethiopian Coffee Producers to International Markets (ACDI. VOCA Technical Bulletin: Addis Ababa, Ethiopia.

134 References

Denu, D., Platts, P. J., Kelbessa, E., Gole, T. W., & Marchant, R. (2016). The role of traditional coffee management in forest conservation and carbon storage in the Jimma Highlands, Ethiopia. Forests, Trees and Livelihoods, 25(4), 226-238. Dercon, S., & Gollin, D. (2014). Agriculture in African development: A review of theories and strategies. Des Marais, D. L., Hernandez, K. M., & Juenger, T. E. (2013). Genotype-by-environment interaction and plasticity: exploring genomic responses of plants to the abiotic environment. Annual Review of Ecology, Evolution, and Systematics, 44, 5-29. Dessalegn, Y., Labuscagne, M., Osthoff, G., & Herselman, L. (2007). Variation of green bean caffeine, chlorogenic acid, sucrose and trigolline contents among Ethiopian Arabica coffee accessories. SINET: Ethiopian Journal of Science, 30(1), 77-82. Dessie, N. (2008). Physical quality standards and grading system of Ethiopian coffee in demand supply chain: In coffee diversity and knowledge. EIAR. AddisAbaba, Ethiopia, 307-317. Devasagayam, T., Kamat, J., Mohan, H., & Kesavan, P. (1996). Caffeine as an antioxidant: inhibition of lipid peroxidation induced by reactive oxygen species. Biochimica et Biophysica Acta (BBA)-Biomembranes, 1282(1), 63-70. Diao, X., Taffesse, A. S., Yu, B., & Pratt, A. N. (2010). Economic importance of agriculture for sustainable development and poverty reduction: The case study of Ethiopia. In Global forum on agriculture, (pp. 29-30). Dominic, S., & Morten, S. (2011). Ethiopian coffee quality improvement project. Addis Ababa, Ethiopia. Dorr, A. C., da Costa, M. L., & dos Reys, M. A. (2010). Análise dos canais de comercialização: o caso dos produtos agroindustrializados no mercado público. Revista Economia & Gestão, 10(23), 8-22. Dorsey, J., & Assefa, T. (2005). Evaluation of Agricultural Cooperatives in Ethiopia (ACE) Program Activities. The Mitchell Group, Inc., Washington, DC. Downey, G., & Boussion, J. (1996). Authentication of Coffee Bean Variety by Near- infrared Reflectance Spectroscopy of Dried Extract. Journal of the Science of Food and Agriculture 71(1), 41-49. Downey, G., Briandet, R., Wilson, R. H., & Kemsley, E. K. (1997). Near-and mid-infrared spectroscopies in food authentication: coffee varietal identification. Journal of Agricultural and Food Chemistry, 45(11), 4357-4361. Downey, G., & Spengler, B. (1996). Compositional analysis of coffee blends by near infrared spectroscopy. Irish Journal of Agricultural and Food Research, 35(2), 179-188. Duarte, G. S., Pereira, A. A., & Farah, A. (2010). Chlorogenic acids and other relevant compounds in Brazilian coffees processed by semi-dry and wet post-harvesting methods. Food chemistry, 118(3), 851-855. Ebrahimi-Najafabadi, H., Leardi, R., Oliveri, P., Chiara Casolino, M., Jalali-Heravi, M., & Lanteri, S. (2012). Detection of addition of barley to coffee using near infrared spectroscopy and chemometric techniques. Talanta, 99, 175-179. Ellis, A. M., & Mayhew, C. A. (2013). Proton transfer reaction mass spectrometry: principles and applications: John Wiley & Sons. Emana, B. (2009). Cooperatives: a path to economic and social empowerment in Ethiopia: ILO. Esteban-Dıez, I., González-Sáiz, J., & Pizarro, C. (2004). An evaluation of orthogonal signal correction methods for the characterisation of arabica and robusta coffee varieties by NIRS. Analytica Chimica Acta, 514(1), 57-67.

135 References

Esteban-Díez, I., González-Sáiz, J. M., Sáenz-González, C., & Pizarro, C. (2007). Coffee varietal differentiation based on near infrared spectroscopy. Talanta, 71(1), 221- 229. Evangelista, S. R., Silva, C. F., da Cruz Miguel, M. G. P., de Souza Cordeiro, C., Pinheiro, A. C. M., Duarte, W. F., & Schwan, R. F. (2014). Improvement of coffee beverage quality by using selected yeasts strains during the fermentation in dry process. Food Research International, 61, 183-195. FAO. (2012). The State of Food Insecurity in the World. Economic growth is necessary but not sufficient to accelerate reduction of hunger and malnutrition. FAO,Rome, Italy. FAO. (2015). The State of Food Insecurity in the World. Meeting the 2015 international hunger targets: taking stock of uneven progress. Rome, FAO. Farah, & Donangelo, C. M. (2006). Phenolic compounds in coffee. Brazilian Journal of Plant Physiology, 18(1), 23-36. Farah, Monteiro, M. C., Calado, V., Franca, A. S., & Trugo, L. C. (2006). Correlation between cup quality and chemical attributes of Brazilian coffee. Food chemistry, 98(2), 373-380. Farah, A. (2012). Coffee constituents. Coffee: Emerging health effects and disease prevention, 1, 22-58. Feldman, R., Ryder, W., & Kung, J. (1969). Importance of nonvolatile compounds to the flavor of coffee. Journal of agricultural and food chemistry, 17(4), 733-739. Feria-Morales, A. M. (2002). Examining the case of green coffee to illustrate the limitations of grading systems/expert tasters in sensory evaluation for quality control. Food Quality and Preference, 13(6), 355-367. Ferreira, E., Pereira, R., Delbem, A., Oliveira, O., & Mattoso, L. (2007). Random subspace method for analysing coffee with electronic tongue. Electronics Letters, 43(21), 1138-1139. Ferreira, M. M., Antunes, A. M., Melgo, M. S., & Volpe, P. L. (1999). Quimiometria I: calibração multivariada, um tutorial. Química Nova, 22(5), 724-731. Field, C. B., Barros, V. R., Mastrandrea, M. D., Mach, K. J., Abdrabo, M.-K., Adger, N., Barnett, J. (2014). Summary for policymakers. Climate change 2014: impacts, adaptation, and vulnerability. Part A: global and sectoral aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, 1-32. Fisk, I. D., Alec, K., Sonja, H., Amarjeet, V., & Javier Silanes, K. (2012). Discrimination of roast and ground coffee aroma. Flavour, 1(14), 1 - 8. Flament, I. (1989). Coffee, cocoa, and tea. Food Reviews International, 5(3), 317-414. Flament, I. (2002). Coffee flavour chemistry. Chichester, UK: John Wiley and Sons. Franca, A. S., Oliveira, L. S., Oliveira, R. C. S., Agresti, P. C. M., & Augusti, R. (2009). A preliminary evaluation of the effect of processing temperature on coffee roasting degree assessment. Journal of Food Engineering, 92(3), 345-352. Fridell, G. (2014). Coffee: John Wiley & Sons. Fukunaga, T., Mori, S., Nakabayashi, Y., Kanda, M., & Toko, K. (1996). Quantification of taste of coffee using sensor with global selectivity. Sensors and Materials, 8(1), 047-056. Garvelink, W., Wedding, K., & Hanson, S. (2012). Smallholder Agriculture: A Critical Factor in Poverty Reduction and Food Security in Africa. In: CSIS. Center for Strategic and International Studies, Washington, DC. Gelaw, F. (2016). Effects of Personalized Transactions and Institutional Interventions on the Performance of Coffee Markets in Ethiopia. PhD thesis.

136 References

Gemech, F., & Struthers, J. (2007). Coffee price volatility in Ethiopia: effects of market reform programmes. Journal of International development, 19(8), 1131-1142. Geremew, T., Abate, D., Landschoot, S., Haesaert, G., & Audenaert, K. (2016). Occurrence of toxigenic fungi and ochratoxin A in Ethiopian coffee for local consumption. Food Control, 69, 65-73. Geromel, C., Ferreira, L. P., Guerreiro, S. M. C., Cavalari, A. A., Pot, D., Pereira, L. F. P., Marraccini, P. (2006). Biochemical and genomic analysis of sucrose metabolism during coffee (Coffea arabica) fruit development. Journal of Experimental Botany, 57(12), 3243-3258. Gil, P., Mittermeier, R., Hoffmann, M., Pilgrim, J., Goettsch-Mittermeier, C., Lamoreux, J., & Da Fonseca, G. (2004). Hotspots revisited. CEMEX, Mexico City. Giovannucci, D., Von Hagen, O., & Wozniak, J. (2014). Corporate social responsibility and the role of voluntary sustainability standards. In Voluntary Standard Systems, (pp. 359-384): Springer. Grote, U., Basu, A. K., & Chau, N. H. (2007). New frontiers in environmental and social labeling: Springer Science & Business Media. Guyot, Lotodé, E., & R Vincent, J. (1988). Analyse qualitative d'un café Coffea canephora var. Robusta en fonction de la maturité, 2: Application de l'analyse statistique multidimensionnelle. Café, Cacao, Thé (Francia) v. 32 (3) p. 229-242. Guyot, B., Gueule, D., Manez, J., Perriot, J., Giron, J., & Villain, L. (1996). Influence de l’altitude et de l’ombrage sur la qualité des cafés arabica. Plantations, recherche, développement, 3(4), 272-280. Harmand, J., Hergoualc'h, K., Miguel, S. d., Dzib, B., Siles, P., & Vaast, P. (2007). Carbon sequestration in coffee agroforestry plantations of Central America. In: 21st International Conference on Coffee Science, Montpellier, France, 11-15 September, 2006., (pp. 1071-1074): Association Scientifique Internationale du Café (ASIC). Herling, A. W., Burger, H.-J., Schwab, D., Hemmerle, H., Below, P., & Schubert, G. (1998). Pharmacodynamic profile of a novel inhibitor of the hepatic glucose-6- phosphatase system. American Journal of Physiology-Gastrointestinal and Liver Physiology, 274(6), G1087-G1093. Huck, C., Guggenbichler, W., & Bonn, G. (2005). Analysis of caffeine, theobromine and theophylline in coffee by near infrared spectroscopy (NIRS) compared to high- performance liquid chromatography (HPLC) coupled to mass spectrometry. Analytica Chimica Acta, 538(1), 195-203. Huck, C. W., Guggenbichler, W., & Bonn, G. K. (2005). Analysis of caffeine, theobromine and theophylline in coffee by near infrared spectroscopy (NIRS) compared to high-performance liquid chromatography (HPLC) coupled to mass spectrometry. Analytica chimica acta, 538(1–2), 195-203. Hundera, K., Aerts, R., Fontaine, A., Van Mechelen, M., Gijbels, P., Honnay, O., & Muys, B. (2013). Effects of Coffee Management Intensity on Composition, Structure, and Regeneration Status of Ethiopian Moist Evergreen Afromontane Forests. Environmental Management, 51(3), 801-809. Hylander, & Nemomissa, S. (2008). Home garden coffee as a repository of epiphyte biodiversity in Ethiopia. Frontiers in Ecology and the Environment, 6(10), 524-528. ICCO. (2014). Middle Africa briefing note, soft commodities: Coffee. In: Ecobank: The Pan African Bank. ICO.(2016). Trade Statistics Tables: http://www.ico.org/trade_statistics.asp. Illy, A., & Viani, R. (2005). Espresso coffee: the science of quality: Academic Press. IPMS. (2007). Improving Productivity and Market Success (IPMS) 2007 Gomma pilot learning Woreda diagnosis and program design In, (pp. 85).

137 References

ITC. (2011). International Trade Centre (ITC). Trends in the Trade of Certified Coffees. ITC.(2012). International Trade Centre (ITC). Niche Markets for Coffee: Specialty, Environment and Social Aspects. Jackels, S., Jackels, C., Vallejos, C., Kleven, S., Rivas, R., & Fraser-Dauphinee, S. (2006). Control of the Coffee Fermentation Process and Quality of Resulting Roasted Coffee: Studies in the Field Laboratory and on Small Farms in Nicaragua During the 2005-06 Harvest. In: 21st International Scientific Colloquium on Coffee–Post-harvest processing and green coffee quality. JARC. (1996). Recommended production technologies for coffee and associated crops. In: Ethiopian Institute of Agricultural Reasearch (EIAR), Addis ababa, Ethiopia. Jena, P. R., Chichaibelu, B. B., Stellmacher, T., & Grote, U. (2012). The impact of coffee certification on small-scale producers’ livelihoods: a case study from the Jimma Zone, Ethiopia. Agricultural economics, 43(4), 429-440. Jeszka-Skowron, M., Zgoła-Grześkowiak, A., & Grześkowiak, T. (2015). Analytical methods applied for the characterization and the determination of bioactive compounds in coffee. European Food Research and Technology, 240(1), 19-31. Jha, S., Bacon, C. M., Philpott, S. M., Méndez, V. E., Läderach, P., & Rice, R. A. (2014). Shade coffee: update on a disappearing refuge for biodiversity. BioScience, 64(5), 416-428. Jiang, L., Ding, Y., Jiang, F., Li, L., & Mo, F. (2014). Electrodeposited nitrogen-doped graphene/carbon nanotubes nanocomposite as enhancer for simultaneous and sensitive voltammetric determination of caffeine and vanillin. Analytica chimica acta, 833, 22-28. Jing, M., Cai, W., & Shao, X. (2010). Quantitative Determination of the Components in Corn and Tobacco Samples by Using Near-Infrared Spectroscopy and Multiblock Partial Least Squares. Analytical Letters, 43(12), 1910-1921. Joët, L., Andréina, Salmona, J., Doulbeau, S., Descroix, F., Bertrand, B., de Kochko, A., & Dussert, S. (2009). Metabolic pathways in tropical dicotyledonous albuminous seeds: Coffea arabica as a case study. New Phytologist, 182(1), 146-162. Joët, L. A., Bertrand, B., & Dussert, S. (2014). Environmental effects on coffee seed biochemical composition and quality attributes: a genomic perspective. In ASIC: The 25th International Conference on Coffee and Science, September 8th–13th, Armenia, Colombia. Association for Science and Information on Coffee (ASIC), Armenia, Colombia, (pp. 70). Joët, L. A., Descroix, F., Doulbeau, S., Bertrand, B., & Dussert, S. (2010). Influence of environmental factors, wet processing and their interactions on the biochemical composition of green Arabica coffee beans. Food chemistry, 118(3), 693-701. John, A. K. (2011). ECX quality operation manual, Ethiopia Commodity Exchange (ECX), Addis Ababa, Ethiopia. Jouanjean, M. A. (2012). Standards, reputation, and trade: evidence from US horticultural import refusals. World Trade Review 11(3): 438–461. Kaplinsky, R., & Morris, M. (2000). A Handbook for Value Chain Research, prepared for International Development Research Centre, Ottawa. Kemsley, E. K., Ruault, S., & Wilson, R. H. (1995). Discrimination between Coffea arabica and Coffea canephora variant robusta beans using infrared spectroscopy. Food chemistry, 54(3), 321-326. Kimemia, J. (2007). Influence of an Albizzia shade tree on soil chemical properties, coffee plant growth and yields in Kenya. In: 21st International Conference on Coffee Science, Montpellier, France, 11-15 September, 2006., (pp.1199-1202): Association Scientifique Internationale du Café (ASIC).

138 References

Kitila, O., Alamerew, S., Kufa, T., & Garedew, W. (2011). Organoleptic Characterization of Some Limu Coffee (Coffea arabica L.) Germplasm at Agaro, Southwestern Ethiopia. International Journal of Agricultural Research, 6, 537-549. Knapp, D. R. (1979). Handbook of analytical derivatization reactions: John Wiley & Sons. Knopp, S., Bytof, G., & Selmar, D. (2006). Influence of processing on the content of sugars in green Arabica coffee beans. European Food Research and Technology, 223(2), 195-201. Kobayashi, Y., Habara, M., Ikezazki, H., Chen, R., Naito, Y., & Toko, K. (2010). Advanced taste sensors based on artificial lipids with global selectivity to basic taste qualities and high correlation to sensory scores. Sensors, 10(4), 3411-3443. Kodama, Y. (2007). New role of cooperatives in Ethiopia: the case of Ethiopian coffee farmers cooperatives. Koshiro, Y., Zheng, X. Q., Wang, M. L., Nagai, C., & Ashihara, H. (2006). Changes in content and biosynthetic activity of caffeine and trigonelline during growth and ripening of Coffea arabica and Coffea canephora fruits. Plant Science, 171(2), 242-250. Kozuma, K., Tsuchiya, S., Kohori, J., Hase, T., & Tokimitsu, I. (2005). Antihypertensive effect of green coffee bean extract on mildly hypertensive subjects. Hypertension Research, 28(9), 711-718. Krähmer, A., Engel, A., Kadow, D., Ali, N., Umaharan, P., Kroh, L. W., & Schulz, H. (2015). Fast and neat–Determination of biochemical quality parameters in cocoa using near infrared spectroscopy. Food chemistry, 181, 152-159. Krähmer, A., Engel, A., Kadow, D., Ali, N., Umaharan, P., Kroh, L. W., & Schulz, H. (2015). Fast and neat – Determination of biochemical quality parameters in cocoa using near infrared spectroscopy. Food chemistry, 181(0), 152-159. Kramer, D., Breitenstein, B., Kleinwächter, M., & Selmar, D. (2010). Stress metabolism in green coffee beans (Coffea arabica L.): expression of dehydrins and accumulation of GABA during drying. Plant and cell physiology, 51(4), 546-553. Kufa, T., & Burkhardt, J. (2011). Stomatal characteristics in Arabica coffee germplasm accessions under contrasting environments at Jimma, southwestern Ethiopia. International Journal of Botany, 7(1), 63-72. Kufa, T., Yilma, A., Shimber, T., Netsere, A., & Taye, E. (2007). Yield performance of Coffea arabica cultivars under different shade trees at Jimma Research Center, Southwest Ethiopia. In:Proceedings of the second International Symposium on Multi-strata Agroforestry Systems with Perennial Crops, consumers and the environment,17-21 September 2007, IUFRO, CIRAD, CATIE and ICRAF,Turrialba, Costa Rica. Ky, Louarn, J., Dussert, S., Guyot, B., Hamon, S., & Noirot, M. (2001). Caffeine, trigonelline, chlorogenic acids and sucrose diversity in wild Coffea arabica L. and C. canephora P. accessions. Food chemistry, 75(2), 223-230. Labouisse, Bellachew, B., Kotecha, S., & Bertrand, B. (2008). Current status of coffee (Coffea arabica L.) genetic resources in Ethiopia: implications for conservation. Genetic Resources and Crop Evolution, 55(7), 1079-1093. Läderach, Oberthür, T., Cook, S., Estrada Iza, M., Pohlan, J. A., Fisher, M., & Rosales Lechuga, R. (2011). Systematic agronomic farm management for improved coffee quality. Field Crops Research, 120(3), 321-329. Lambert, J. D., Hong, J., Yang, G.-y., Liao, J., & Yang, C. S. (2005). Inhibition of carcinogenesis by polyphenols: evidence from laboratory investigations. The American Journal of Clinical Nutrition, 81(1), 284S-291S. Lara-Estrada, L., & Vaast, P. (2007). Effects of altitude, shade, yield and fertilization on coffee quality (Coffea arabica L. var. Caturra) produced in agroforestry systems of the Northern Central Zones of Nicaragua. In Second International Symposium on

139 References

Multi-Strata Agroforestry Systems with Perennial Crops: Making Ecosystem Services Count for Farmers, Consumers and the Environment, Turrialba, Costa Rica, (pp. 17-21).

Larsson, S. C., Virtamo, J., & Wolk, A. (2011). Coffee consumption and risk of stroke in women. Stroke, 42(4), 908-912. Lashermes, P., Combes, M.-C., Ansaldi, C., Gichuru, E., & Noir, S. (2011). Analysis of alien introgression in coffee tree (Coffea arabica L.). Molecular breeding, 27(2), 223-232. Lashermes, P., Combes, M.-C., Robert, J., Trouslot, P., D'hont, A., Anthony, F., & Charrier, A. (1999). Molecular characterisation and origin of the Coffea arabica L. genome. Molecular and General Genetics MGG, 261(2), 259-266. Lee, C. (2000). Antioxidant ability of caffeine and its metabolites based on the study of oxygen radical absorbing capacity and inhibition of LDL peroxidation. Clinica Chimica Acta, 295(1), 141-154. Lee, L. W., Cheong, M. W., Curran, P., Yu, B., & Liu, S. Q. (2015). Coffee fermentation and flavor–An intricate and delicate relationship. Food chemistry, 185, 182-191 Leloup, V., Gancel, C., Liardon, R., Rytz, A., & Pithon, A. (2005). Impact of wet and dry process on green coffee composition and sensory characteristics. In: ASIC 2004. 20th International Conference on Coffee Science, Bangalore, India,11-15 October 2004.,(pp.93-101): Association Scientifique Internationale du Café (ASIC). Lemma, B. (2015). ECX Adopts Five Level Coffee Quality Grading System, Ethiopia Commodity Exchange (ECX), Addis Ababa, Ethiopia. Leroy, Ribeyre F., Bertrand B., C. P., , D. M., Montagnon C., Marraccini P., & D, P. (2006 ). Genetics of coffee quality. Brazilian Journal of Plant Physiology 18, 229- 242.

Link, J. V., Lemes, A. L. G., Marquetti, I., dos Santos Scholz, M. B., & Bona, E. (2014). Geographical and genotypic segmentation of arabica coffee using self-organizing maps. Food Research International, 59, 1-7. LMC. (2006). Enhancement of Coffee Quality Socioeconomic Study. Madihah, K. K., Zaibunnisa, A., Norashikin, S., Rozita, O., & Misnawi, J. (2013). Optimization of roasting conditions for high-quality Arabica coffee. International Food Research Journal, 20(4). Makri, E., Tsimogiannis, D., Dermesonluoglu, E. K., & Taoukisa, P. S. (2011). Modeling of Greek coffee aroma loss during storage at different temperatures and water activities. Procedia Food Science, 1, 1111-1117. Masoud, W., Bjørg Cesar, L., Jespersen, L., & Jakobsen, M. (2004). Yeast involved in fermentation of Coffea arabica in East Africa determined by genotyping and by direct denaturating gradient gel electrophoresis. Yeast, 21(7), 549-556. Mateos-Aparicio, G. (2011). Partial Least Squares (PLS) Methods: Origins, Evolution, and Application to Social Sciences. Communications in Statistics - Theory and Methods, 40(13), 2305-2317. Matos, A., & Monaco, L. (2001). PA; Pinto, AB; Fia, R.; Fukunaga, DC Pollutant potential of wastewater from the coffee fruits processing. Soares, AA; Saturnino, HM Environmental water: Competitive use and conservation strategies for water and natural resources. Brasília: ABID, 158-165. Mazzafera, P. (1999). Chemical composition of defective coffee beans. Food chemistry, 64(4), 547-554. Mazzafera, P., & Gonçalves, K. V. (1999). Nitrogen compounds in the xylem sap of coffee. Phytochemistry, 50(3), 383-386.

140 References

Mburu, J., Thuo, J., & Marder, R. (1994). The characterization of wastewater from coffee processing factories in Kenya. Kenya Coffee, 59(690), 1757-1762. Meijerink, G., Alemu, D., Mheen, J., & Wijnands, J. (2010). Institutional innovation in African markets: can commodity exchanges address sustainability issues? Mekuria, T., Neuhoff, D., & Kopke, U. (2004). The status of coffee production and the potential for organic conversion in Ethiopia. In: Conference on International Agricultural Research for Development, Berlin. Mendez, V. E., Bacon, C. M., Olson, M., Petchers, S., Herrador, D., Carranza, C., . . . Mendoza, A. (2010). Effects of Fair Trade and organic certifications on small- scale coffee farmer households in Central America and Mexico. Renewable Agriculture and Food Systems, 25(03), 236-251. MFA. (2016). The 4th World Coffee Conference due to unfold in Addis Ababa, Ministry of Foreign Affairs (MFA) of Ethiopia, Addis ababa, Ethiopia. Michishita, T., Akiyama, M., Hirano, Y., Ikeda, M., Sagara, Y., & Araki, T. (2010). Gas chromatography/olfactometry and electronic nose analyses of retronasal aroma of espresso and correlation with sensory evaluation by an artificial neural network. Journal of Food Science, 75(9), S477-S489. Muleta, D., Assefa, F., Nemomissa, S., & Granhall, U. (2008). Distribution of arbuscular mycorrhizal fungi spores in soils of smallholder agroforestry and monocultural coffee systems in southwestern Ethiopia. Biology and Fertility of Soils, 44(4), 653- 659. Murthy, P. S., & Naidu, M. M. (2012). Sustainable management of coffee industry by- products and value addition: A review. Resources, Conservation and recycling, 66, 45-58. Muschler, R. G. (2001). Shade improves coffee quality in a sub-optimal coffee-zone of Costa Rica. Agroforestry systems, 51(2), 131-139. Nehlig, A. (2012). Effets physiologiques du café et santé humaine. Une revue. Cahiers Agricultures, 21(2), 197-207. Niggeweg, R., Michael, A. J., & Martin, C. (2004). Engineering plants with increased levels of the antioxidant chlorogenic acid. Nature biotechnology, 22(6), 746-754. O'Keefe, J. H., Bhatti, S. K., Patil, H. R., DiNicolantonio, J. J., Lucan, S. C., & Lavie, C. J. (2013). Effects of habitual coffee consumption on cardiometabolic disease, cardiovascular health, and all-cause mortality. Journal of the American College of Cardiology, 62(12), 1043-1051. Oberthür, T., Läderach, P., H.A Jurgen, P., & cock, J. H. (2012). specialty coffee: Managing quality. In): Internation plant nutrition Institute, southeast Asia Program. Oberthür, T., Läderach, P., Posada, H., Fisher, M. J., Samper, L. F., Illera, J., Villegas, A. (2011). Regional relationships between inherent coffee quality and growing environment for denomination of origin labels in Nariño and Cauca, Colombia. Food Policy, 36(6), 783-794. Villegas, A. (2011). Regional relationships between inherent coffee quality and growing environment for denomination of origin labels in Nariño and Cauca, Colombia. Food Policy, 36(6), 783-794. Ochiai, R., Jokura, H., Suzuki, A., Tokimitsu, I., Ohishi, M., Komai, N., . . . Ogihara, T. (2004). Green coffee bean extract improves human vasoreactivity. Hypertension Research, 27(10), 731-737. Oestreich-Janzen, S. (2010). Chemistry of coffee. Comprehensive natural products II, 1085-1117. Olthof, M. R., Hollman, P. C., & Katan, M. B. (2001). Chlorogenic acid and caffeic acid are absorbed in humans. The Journal of Nutrition, 131(1), 66-71.

141 References

Ongo, E., Falasconi, M., Sberveglieri, G., Antonelli, A., Montevecchi, G., Sberveglieri, V., Sevilla III, F. (2012). Chemometric discrimination of Philippine civet coffee using electronic nose and gas chromatography mass spectrometry. Procedia Engineering, 47, 977-980. Parras, P., Martínez-Tomé, M., Jiménez, A., & Murcia, M. (2007). Antioxidant capacity of coffees of several origins brewed following three different procedures. Food chemistry, 102(3), 582-592. Pearl, H., Nagai, C., Moore, P., Steiger, D., Osgood, R., & Ming, R. (2004). Construction of a genetic map for arabica coffee. Theoretical and Applied Genetics, 108(5), 829-835. Pendergrast, M. (2009). Coffee: second to oil? The tea & coffee trade journal, 181(4), 38. Perfecto, I., Vandermeer, J., Mas, A., & Pinto, L. S. (2005). Biodiversity, yield, and shade coffee certification. Ecological Economics, 54(4), 435-446. Perrois, C., Strickler, S. R., Mathieu, G., Lepelley, M., Bedon, L., Michaux, S,. Privat, I. (2015). Differential regulation of caffeine metabolism in Coffea arabica (Arabica) and Coffea canephora (Robusta). Planta, 241(1), 179-191. Perrone, D., Donangelo, C. M., & Farah, A. (2008). Fast simultaneous analysis of caffeine, trigonelline, nicotinic acid and sucrose in coffee by liquid chromatography–mass spectrometry. Food chemistry, 110(4), 1030-1035. Pesta, D. H., Angadi, S. S., Burtscher, M., & Roberts, C. K. (2013). The effects of caffeine, nicotine, ethanol, and tetrahydrocannabinol on exercise performance. Nutrition & metabolism, 10(1), 1. Peterson, J., Harrison Jr, H., Snook, M., & Jackson, D. (2005). Content of chlorogenic acid in sweetpotato germplasm and possible roles in disease and pest resistance. Allelopathy J, 16, 239-250. Petit, N. (2007). Ethiopia’s Coffee Sector: A Bitter or Better Future? Journal of Agrarian Change, 7( 2), 225 - 263. Philpott, S. M., Bichier, P., Rice, R., & Greenberg, R. (2007). Field-Testing Ecological and Economic Benefits of Coffee Certification Programs. Conservation Biology, 21(4), 975-985. Pierrot, J., Giovannucci, D., & Kasterine, A. (2010). Trends in the trade of certified coffees. International Trade Centre Technical Paper. Pimenta, C. J. (2003). Qualidade de café: Editora UFLA. Pizarro, C., Esteban-Dı́ez, I., Nistal, A.-J., & González-Sáiz, J.-M. a. (2004). Influence of data pre-processing on the quantitative determination of the ash content and lipids in roasted coffee by near infrared spectroscopy. Analytica chimica acta, 509(2), 217-227. Poncelet, M., Defourny, J., & De Pelsmaker, P. (2005). A Fair and Sustainable Trade Between Market and Solidarity: Diagnosis and Perspectives, Brussels, Services Scientifiques Techniques et Culturels (SSTC), Ministère de la Recherche Scientifique-Gouvernement Fédéral Belge, January 2005, 100p. Belgian science policy. Ponte, S. (2002). The 'latte revolution'? Regulation, markets and consumption in the global coffee chain. World development, 30(7), 1099-1122. Potts, J., Lynch, M., Wilkings, A., Huppé, G., Cunningham, M., & Voora, V. (2014). The state of sustainability initiatives review 2014: Standards and the green economy. International Institute for Sustainable Development (IISD) and the International Institute for Environment and Development (IIED), 332. Prakash, N. S., Combes, M.-C., Somanna, N., & Lashermes, P. (2002). AFLP analysis of introgression in coffee cultivars (Coffea arabica L.) derived from a natural interspecific hybrid. Euphytica, 124(3), 265-271.

142 References

Privat, I., Foucrier, S., Prins, A., Epalle, T., Eychenne, M., Kandalaft, L., Foyer, C. (2008). Differential regulation of grain sucrose accumulation and metabolism in Coffea arabica (Arabica) and Coffea canephora (Robusta) revealed through gene expression and enzyme activity analysis. New Phytologist, 178(4), 781-797. Puerta Quintero, G. (2001). Strategies to guarantee the quality of the beverage in Colombian CoffeesProceedings. In: 19th. International Scientific Colloquium on Coffee14-18 May 2001Trieste (Italia)): Association Scientifique Internationale du Cafe, París (Francia). Raina, S., Mukai, Y., & Yamamoto, M. (1998). In situ hybridization identifies the diploid progenitor species of Coffea arabica (Rubiaceae). Theoretical and Applied Genetics, 97(8), 1204-1209. Reis, N., Franca, A. S., & Oliveira, L. S. (2013). Discrimination between roasted coffee, roasted corn and coffee husks by Diffuse Reflectance Infrared Fourier Transform Spectroscopy. LWT - Food Science and Technology, 50(2), 715-722. Ribeiro, J., Augusto, F., Salva, T., Thomaziello, R., & Ferreira, M. (2009). Prediction of sensory properties of Brazilian Arabica roasted coffees by headspace solid phase microextraction-gas chromatography and partial least squares. Analytica Chimica Acta, 634(2), 172-179. Ribeiro, J. S., Augusto, F., Salva, T. J. G., & Ferreira, M. M. C. (2012). Prediction models for Arabica coffee beverage quality based on aroma analyses and chemometrics. Talanta, 101(0), 253-260. Ribeiro, J. S., Augusto, F., Salva, T. J. G., Thomaziello, R. A., & Ferreira, M. M. C. (2009). Prediction of sensory properties of Brazilian Arabica roasted coffees by headspace solid phase microextraction-gas chromatography and partial least squares. Analytica chimica acta, 634(2), 172-179. Ribeiro, J. S., Ferreira, M. M. C., & Salva, T. J. G. (2011). Chemometric models for the quantitative descriptive sensory analysis of Arabica coffee beverages using near infrared spectroscopy. Talanta, 83(5), 1352-1358. Robert, S. (2015). Frontlines: Foreign aid impact in US and Abroad, Washington, DC, USA. In Frontlines, (pp. 17): USID. Robinson, W. E., Reinecke, M. G., Abdel-Malek, S., Jia, Q., & Chow, S. A. (1996). Inhibitors of HIV-1 replication that inhibit HIV integrase. Proceedings of the National Academy of Sciences, 93(13), 6326-6331. Roche, D. (1995). Coffee genetics and quality. In:Colloque Scientifique International sur le Café, 16. Kyoto (Japón), Avril 9-14. Rodriguez-Saona, L. E., & Allendorf, M. E. (2011). Use of FTIR for Rapid Authentication and Detection of Adulteration of Food. Annual Review of Food Science and Technology, 2(1), 467-483. Rodríguez, J., Durán, C., & Reyes, A. (2009). Electronic nose for quality control of Colombian coffee through the detection of defects in “Cup Tests”. Sensors, 10(1), 36-46. Ruben, R., & Fort, R. (2012). The impact of fair trade certification for coffee farmers in Peru. World development, 40(3), 570-582. Salazar-Martinez, E., Willett, W. C., Ascherio, A., Manson, J. E., Leitzmann, M. F., Stampfer, M. J., & Hu, F. B. (2004). Coffee consumption and risk for type 2 diabetes mellitus. Annals of internal medicine, 140(1), 1-8. Santos, J. R., Sarraguça, M. C., Rangel, A. O., & Lopes, J. A. (2012). Evaluation of green coffee beans quality using near infrared spectroscopy: A quantitative approach. Food chemistry, 135(3), 1828-1835.

143 References

Santos, K. M., Moura, M. F., Azevedo, F. G., Lima, K. M., Raimundo Jr, I. M., & Pasquini, C. (2012). Classification of Brazilian coffee using near-infrared spectroscopy and multivariate calibration. Analytical Letters, 45(7), 774-781. Sberveglieri, G., Concina, I., Comini, E., Falasconi, M., Ferroni, M., & Sberveglieri, V. (2012). Synthesis and integration of tin oxide nanowires into an electronic nose. Vacuum, 86(5), 532-535. Schenker, S. (2000). Investigations on the hot air roasting of coffee beans. Diss. Technische Wissenschaften ETH Zürich, Nr. 13620, 2000. Schwan, R., & Wheals, A. (2003). Mixed microbial fermentations of chocolate and coffee. Yeasts in Food. Hamburg, Alemanha: Behrs Verlag, 1, 426-459. Selmar, Bytof, G., & Knopp, S. (2002). New aspects of coffee processing: the relation between seed germination and coffee quality. In:Proceeding of the "19th International Coffee Science Conference ". ASIC, Paris.Selmar, Bytof, G., Knopp, S., & Breitenstein, B. (2006). Germination of coffee seeds and its significance for coffee quality. Plant Biology, 8(2), 260-264. Selmar, Bytof, G., Knopp, S. E., & Breitenstein, B. (2006). Germination of coffee seeds and its significance for coffee quality. Plant Biology, 8(2), 260-264. Silva, C. F., Vilela, D. M., de Souza Cordeiro, C., Duarte, W. F., Dias, D. R., & Schwan, R. F. (2013). Evaluation of a potential starter culture for enhance quality of coffee fermentation. World Journal of Microbiology and Biotechnology, 29(2), 235-247. Silva, E. A. d., Mazzafera, P., Brunini, O., Sakai, E., Arruda, F. B., Mattoso, L. H. C., Pires, R. C. M. (2005). The influence of water management and environmental conditions on the chemical composition and beverage quality of coffee beans. Brazilian Journal of Plant Physiology, 17(2), 229-238. Silvarolla, M. B., Mazzafera, P., & Fazuoli, L. C. (2004). Plant biochemistry: a naturally decaffeinated arabica coffee. Nature, 429(6994), 826-826. Silvarolla, M. B., Mazzafera, P., & Lima, M. M. A. d. (2000). Caffeine content of Ethiopian Coffea arabica beans. Genetics and Molecular Biology, 23(1), 213-215. Singham, P., Birwal, P., & Yadav, B. (2015). Importance of Objective and Subjective Measurement of Food Quality and their Inter-relationship. Journal of Food Processing & Technology, 6(9), 1. Slagle, A., Skousen, J., Bhumbla, D., Sencindiver, J., & McDonald, L. (2004). Trace element concentrations of three soils in central Appalachia. Soil Horizons, 45(3), 73-85. Smith, A. W. (1985). Agricultural practices. In R. J. Clarke & R. Macrae (Eds.) Coffee chemistry,Elsevier, Applied Science: Amsterdam, 18–23. Soto-Pinto, L., Perfecto, I., Castillo-Hernandez, J., & Caballero-Nieto, J. (2000). Shade effect on coffee production at the northern Tzeltal zone of the state of Chiapas, Mexico. Agriculture, Ecosystems & Environment, 80(1), 61-69. Sridevi, V., & Parvatam, G. (2014). Changes in caffeine content during fruit development in Coffea canephora P. ex. Fr. grown at different elevations. Journal of Biology and Earth Sciences, 4(2), B168-B175. Stalmach, A., Steiling, H., Williamson, G., & Crozier, A. (2010). Bioavailability of chlorogenic acids following acute ingestion of coffee by humans with an ileostomy. Archives of biochemistry and biophysics, 501(1), 98-105. Stellmacher, T. (2007). Governing the Ethiopian coffee forests: a local level institutional analysis in Kaffa and Bale Mountains: Shaker. Stellmacher, T. (2011). Forest Coffee Certification in Ethiopia: Economic Boon or Ecological Bane? Stich, H., Rosin, M., & Bryson, L. (1982). Inhibition of mutagenicity of a model nitrosation reaction by naturally occurring phenolics, coffee and tea. Mutation

144 References

Research/Fundamental and Molecular Mechanisms of Mutagenesis, 95(2), 119- 128. Subedi, R. N. (2011). Comparative analysis of dry and wet processing of coffee with respect to quality and cost in Kavre District, Nepal: A case of Panchkhal Village. International Research Journal of Applied and Basic Sciences, 2(5), 181-193. Tafesse, G. (2005). The contributions of agricultural growth to poverty reduction in Ethiopia. Ethiopian Journal of Economics, 14(1), 1-25. Tefera, A. (2015). Ethiopia: Coffee Annual Report, GAIN Report ET-1514, USDA. Tefera, A. (2016). Ethiopia: Coffee Annual report: Coffee Production and Exports Remain Steady. Tefera, A., & Tefera, T. (2014). Ethiopia: Coffee Annual Report, GAIN Report ET-1402, USDA

Thom, E. (2007). The effect of chlorogenic acid enriched coffee on glucose absorption in healthy volunteers and its effect on body mass when used long-term in overweight and obese people. Journal of International Medical Research, 35(6), 900-908.

Toci, A., & Farah, A. (2010). Contents of caffeine, trigonelline, chlorogenic acids and sucrose in brazilian defective beans. In: Proceedings of the 23th International Colloquium on the Chemistry of Coffee. Toci, A. T., & Farah, A. (2008). Volatile compounds as potential defective coffee beans’ markers. Food chemistry, 108(3), 1133-1141. Toledo, P. R., Pezza, L., Pezza, H. R., & Toci, A. T. (2016). Relationship Between the Different Aspects Related to Coffee Quality and Their Volatile Compounds. Comprehensive Reviews in Food Science and Food Safety. Tressl, R., Holzer, M., & Kamperschroer, H. (1982). Buidung von Aromastaffen in Rostkaffee in Abhangigkeit vom Gehalt an Aminosauren und Reduzierenden Zuckern. In: 10th International Colloquium on the chemistry of coffee, (pp. 111- 114). Trugo, & Macrae, R. (1984). Chlorogenic acid composition of instant coffees. Analyst, 109(3), 263-266. Trugo, & Macrae, R. (1984). A study of the effect of roasting on the chlorogenic acid composition of coffee using HPLC. Food chemistry, 15(3), 219-227. Tsegaye, B., Mohammed, A., Shimber, T., Getachew, Y., & Getachew, E. (2014). The Influence of sun drying methods and layer thickness on the quality of lowland Arabica coffee varieties at Gomma I, Southewesten Ethiopia. Research Journal of Agriculture and Environmental Management. Vol, 3(11), 547-554. Tunson, J. (2014). Effects of Habitual Coffee Consumption on Cardiometabolic Disease, Cardiovascular Health, and All-cause Mortality. Journal of Emergency Medicine, 46(5), 749. Upadhyay, R., & Mohan Rao, L. J. (2013). An outlook on chlorogenic acids— occurrence, chemistry, technology, and biological activities. Critical reviews in food science and nutrition, 53(9), 968-984. Vaast, & Bertrand, B. (2005). Date of harvest and altitude influence bean characteristics and beverage quality of Coffea arabica in intensive management conditions. In): HortScience. Vaast, Bertrand, B., Perriot, J. J., Guyot, B., & Génard, M. (2006). Fruit thinning and shade improve bean characteristics and beverage quality of coffee (Coffea arabica L.) under optimal conditions. Journal of the Science of Food and Agriculture, 86(2), 197-204.

145 References

Vaast, Cilas, C., Perriot, J., Davrieux, F., Guyot, B., & Bolano, M. (2005a). Mapping of coffee quality in Nicaragua according to regions, ecological conditions and farm management. In: ASIC 2004. 20th International Conference on Coffee Science, Bangalore, India, 11-15 October 2004., (pp. 842-850): Association Scientifique Internationale du Café (ASIC). Vaast, Kanten, R. v., Siles, P., Dzib, B., Franck, N., Harmand, J., & Genard, M. (2005). Shade: a key factor for coffee sustainability and quality. In ASIC 2004. 20th International Conference on Coffee Science, Bangalore, India, 11-15 October 2004., (pp. 887-896): Association Scientifique Internationale du Café (ASIC). Van Dam, R. M., & Hu, F. B. (2005). Coffee consumption and risk of type 2 diabetes: a systematic review. Jama, 294(1), 97-104. Van der Stegen, G., & Van Duijn, J. (1980). Analysis of chlorogenic acids in coffee. In Colloque Scientifique International sur le Café, 9. Londres (Inglaterra), Juin 16-20, 1980. Van der Vossen, H. A. (1985). Coffee selection and breeding. In Coffee, (pp. 48-96): Springer. Van Oijen, M., Dauzat, J., Harmand, J.-M., Lawson, G., & Vaast, P. (2010). Coffee agroforestry systems in Central America: I. A review of quantitative information on physiological and ecological processes. Agroforestry systems, 80(3), 341-359. Van Oirschot, Q. E., & Tomlins, K. I. (2002). Applying analytical sensory evaluation techniques, which translate qualitative perceptions to numerical data to research on development issues. In: Conference on Combining Qualitative and Quantitative Methods in Development Research, Centre for Development Studies, University of Wales, Swansea, (pp. 1-2): Citeseer. Vignoli, J., Bassoli, D., & Benassi, M. (2011). Antioxidant activity, polyphenols, caffeine and melanoidins in soluble coffee: The influence of processing conditions and raw material. Food chemistry, 124(3), 863-868. Villarreal, D., Laffargue, A., Posada, H., Bertrand, B., Lashermes, P., & Dussert, S. (2009). Genotypic and environmental effects on coffee (Coffea arabica L.) bean fatty acid profile: impact on variety and origin chemometric determination. Journal of agricultural and food chemistry, 57(23), 11321-11327 Vitoria, A., & Mazzafera, P. (1997). Cytokinin-like effects of caffeine in bioassays. Biologia Plantarum, 40(3), 329-335. Vlasov, Y., & Legin, A. (1998). Non-selective chemical sensors in analytical chemistry: from “electronic nose” to “electronic tongue”. Fresenius' journal of analytical chemistry, 361(3), 255-260. Volkmann, J. (2008). Assessment of certification options for wild forest coffee in Ethiopia. In: Coffice Project Report Sub-project. Von Enden, J. C., Calvert, K. C., Sanh, K., Hoa, H., Tri, Q., & Vietnam, S. (2002). Review of coffee waste water characteristics and approaches to treatment. Project," Improvement of Coffee Quality and Sustainability of Coffee Production in Vietnam". German Technical Cooperation Agency (GTZ), 1-10. Walyaro, D. (1983). Considerations in breeding for improved yield and quality in arabica coffee (Coffea arabica L.). Landbouwhogeschool te Wageningen. Wang, G., Bhoopalan, V., Wang, D., Wang, L., & Xu, X. (2015). The effect of caffeine on cisplatin-induced apoptosis of lung cancer cells. Experimental hematology & oncology, 4(1), 1. Wang, J., Jun, S., Bittenbender, H. C., Gautz, L., & Li, Q. X. (2009). Fourier Transform Infrared Spectroscopy for Kona Coffee Authentication. Journal of Food Science, 74(5), C385-C391.

146 References

Wei, F., Furihata, K., Miyakawa, T., Koda, M., Hu, F., & Tanokura, M. (2013). Comprehensive analysis of coffee bean extracts by NMR spectroscopy: time course of changes in composition. Magnetic Resonance in Food Science: Food for Thought, 343, 183. Wellman, F. L. (1961). Coffee: botany, cultivation, and utilization. Coffee: Botany, Cultivation, and Utilization. Wilson, A. D., & Baietto, M. (2009). Applications and advances in electronic-nose technologies. Sensors, 9(7), 5099-5148. Wintgens, J. N. (2004). The coffee plant. Coffee: Growing, Processing, Sustainable Production: A Guidebook for Growers, Processors, Traders, and Researchers,Wiley-VCH, Weinheim,Germany, pp 1-24. Wintgens, J. N. (2012). Coffee: Growing, processing, sustainable production. A guidebook for growers, processors, traders and researchers (2nd ed.). Weinheim, Germany Wood. (2015). Research and Markets: Coffee Market in Brazil 2015-2019 - High Demand for Arabica Coffee , Technavio, Elmhurst, USA Worku, M., Duchateau, L., & Boeckx, P. (2016). Reproducibility of coffee quality cupping scores delivered by cupping centers in Ethiopia. Journal of Sensory Studies, 31(5), 423-429. World Bank. (2006). Ethiopia: Policies for Pro-Poor Agricultural Growth. Washington, DC: World Bank. EDRI (Ethiopian Development Research Institute). Woubie, A. A., Muradian, R., & Ruben, R. (2015). Impact of multiple certification on smallholder coffee farmers’ livelihoods: evidence from southern Ethiopia. Coffee certification in East Africa, 127. Wrigley, G. (1988). Coffee: Longman Scientific & Technical New York. Xie, L., Ye, X., Liu, D., & Ying, Y. (2009). Quantification of glucose, fructose and sucrose in bayberry juice by NIR and PLS. Food chemistry, 114(3), 1135-1140. Yadessa, A., Burkhardt, J., Denich, M., Woldemariam, T., Bekele, E., & Goldbach, H. (2009). Effect of different indigenous shade trees on the quality of wild arabica coffee in the Afromontane rainforests of Ethiopia. In: 22nd International Conference on Coffee Science, ASIC 2008, Campinas, SP, Brazil, 14-19 September, 2008., (pp. 1227-1233): Association Scientifique Internationale du Café (ASIC). Yeretzian, C., Jordan, A., & Lindinger, W. (2003). Analysing the headspace of coffee by proton-transfer-reaction mass-spectrometry. International Journal of Mass Spectrometry, 223, 115-139. You, D.-C., Kim, Y.-S., Ha, A.-W., Lee, Y.-N., Kim, S.-M., Kim, C.-H., Lee, J.-M. (2011). Possible health effects of caffeinated coffee consumption on Alzheimer's disease and cardiovascular disease. Toxicological research, 27(1), 7. Yukawa, G., Mune, M., Otani, H., Tone, Y., Liang, X.-M., Iwahashi, H., & Sakamoto, W. (2004). Effects of coffee consumption on oxidative susceptibility of low-density lipoproteins and serum lipid levels in humans. Biochemistry (Moscow), 69(1), 70- 74. Zhang, X., Li, W., Yin, B., Chen, W., Kelly, D. P., Wang, X., Du, Y. (2013). Improvement of near infrared spectroscopic (NIRS) analysis of caffeine in roasted Arabica coffee by variable selection method of stability competitive adaptive reweighted sampling (SCARS). Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 114, 350-356. Zhao, Y., Wang, J., Ballevre, O., Luo, H., & Zhang, W. (2012). Antihypertensive effects and mechanisms of chlorogenic acids. Hypertension Research, 35(4), 370-374.

147

Appendix Ethiopia Commodity Exchange Preliminary Unwashed Coffee Quality Assessment Raw Value Moisture Content ______%Date Cup Value Retained on Screen 14 ______%Code No Total Points Origin Grading Table RAW VALUE 40% CUP VALUE 60% Cup Cleanness Defects (30%) Odor (10%) Acidity (15%) Body (15%) Flavour (15%) (15%) Primary Pts Secndary PtsQuality Pts Quality Pts Intensity Pts Quality Pts (count) (Weight) Quality Pts (15%) (15%) <5 15 <5 15Clean 10 Clean 15 Pointed 15 Full 15 Good 15 6-10 12 <10 12F. Clean 8 F. clean 12 M.pointed 12 M. full 12 F. good 12 11-15 9 <15 9Trace 6 1 CD 9 Medium 9 Medium 9 Average 9 16-20 6 <20 6Light 4 2 CD 6 Light 6 Light 6 Fair 6 21-25 3 <25 3Moderate 2 3 CD 3 Lacking 3 Thin 3 Commonish 3 >25 1.5 >25 1.5Strong 0 >3 CD 0 N.D 0 N.D 0 N.D 0 Classification Grade & Points Raw Defects Yirgachefe A UYCA Grade 1= 91-100 SCAA Primary Defects Secondary Defects Observations Yirgachefe B UYCB Grade 2= 81-90 Type Bean Grad SCAA 0 1 2 3 Ethiopia 0 1 2 3 Jimma A UJMA Grade 3= 71-80Full Black Partial Black Foxy Jimma B UJMB Grade 4= 63-70 Full Sour Partial Sour Under Dried Sidama A USDA Grade 5= 58-62Fungus Floater Over Dried Sidama B USDB Grade 6= 50-57 F. Matter Immature Mixed Drying Sidama C USDC Grade 7= 40-49 Insect D. Withered Stinkers Harar A UHRA Grade 8= 31-39 Pod/Husk Shell Faded Harar B UHRB Grade 9= 20-30 S. Insect D. Coated Harar C UHRC UG= 15-19 Broken Light Harar D UHRD Total Primary Soiled Starved Nekempti ULK Grade = (Transfer Forest A UFRA to Grading Table) Total Forest B UFRB

Classification Name Signature Grade Coordinator: C.D- Cup Defect Cupper 1: N.D- Not Detected Cupper 2;

149

Appendix

Ethiopia Commodity Exchange Semi-Washed Coffee Quality Assessment

Raw Value ______Moisture Content ______%Date______Cup Value ______Retained on Screen 14______%Code No______Total Points ______Origin ______

Grading Table RAW VALUE 40% CUP VALUE 60% Cup Cleanness Defects (20%) Shape & Make (10%) Color (5%) Odor (5%) Acidity (15%) Body (15%) Flavour (15%) (15%) Primary Pts Secondar PtsQuality PtsQuality PtsQuality Pts Quality Pts Intensity Pts Quality Pts (count) y (weight) (10%) (10%) Quality Pts 0 10 <5% 10V. good 10Bluish 5Clean 5 Clean 15 Pointed 15 Full 15 Good 15 1-4 8 <8% 8Good 8Grayish 4F. Clean 4 F. clean 12 M.pointed 12 M. full 12 F. good 12 5-6 6 <10% 6F. good 6Greenish 3Trace 3 1 CD 9 Medium 9 Medium 9 Average 9 7-10 4 <12% 4Average 4Coated 2Light 2 2 CD 6 Light 6 Light 6 Fair 6 11-15 2 <14% 2Fair 2Faded 1Moderate 1 3 CD 3 Lacking 3 Thin 3 Commonish 3 >15 1 >14% 1Small 1White 0Strong 0 >3 CD 0 N.D 0 N.D 0 N.D 0

Classification Grade & Points Parchment Observations Raw Defects 0123 0123SCAA Primary Defects Secondary Defects Observations Type Bean Grade SCAA 0 1 2 3 Ethiopia 0 1 2 3 Yirgachefee A WYCA Grade 1= 91-100 Even Under Full Black Partial Foxy Yirgachefee B WYCB Grade 2= 81-90 Undergra Cracked Full Sour Partial Under Sidama A WSDA Grade 3= 71-80 Improper Dull Fungus Floater Over Dried Sidama B WSDB Grade 4= 63-70 Discolore Pods F. Matter Immature Mixed Sidama C WSDC Grade 5= 58-62 Nipped Mixed Insect D. Withered Stinkers Limu A WLMA Grade 6= 50-57 Fermente Under Shell Faded Limu B WLMB Grade 7= 40-49 Loose Long Cont. S. Insect Coated Tepi WTP Grade 8= 31-39 Brownish Total Primary Light Grade = (Transfer Bebeka WBB Grade 9= 20-30 Total Starved to Grading Table) Nekempti WLK UG= 15-19

Classification Name Signature

Grade Coordinator:

Cupper 1:

Cupper 2;

150

Appendix

Ethiopia Commodity Exchange Preliminary Washed Coffee Quality Assessment

Raw Value ______Moisture Content ______%Date______Cup Value ______Retained on Screen 14______%Code No______Total Points ______Origin ______

Grading Table RAW VALUE 40% CUP VALUE 60% Cup Cleanness Defects (20%) Shape & Make (10%) Color (5%) Odor (5%) Acidity (15%) Body (15%) Flavour (15%) (15%) Primary Pts Secondar PtsQuality PtsQuality PtsQuality Pts QualityPts Intensity Pts Quality Pts (count) y (weight) (10%) (10%) Quality Pts 0 10 <5% 10V. good 10Bluish 5Clean 5 Clean 15 Pointed 15 Full 15 Good 15 1-4 8 <8% 8Good 8Grayish 4F. Clean 4 F. clean12 M.pointed 12 M. full 12 F. good 12 5-6 6 <10% 6F. good 6Greenish 3Trace 3 1 CD 9 Medium 9 Medium 9 Average 9 7-10 4 <12% 4Average 4Coated 2Light 2 2 CD 6 Light 6 Light 6 Fair 6 11-15 2 <14% 2Fair 2Faded 1Moderate 1 3 CD 3 Lacking 3 Thin 3 Commonish 3 >15 1 >14% 1Small 1White 0Strong 0 >3 CD 0 N.D 0 N.D 0 N.D 0

Classification Name Signature

Grade Coordinator:

C.D- Cup Defect Cupper 1:

N.D- Not Detected Cupper 2;

151

Appendix

ECX Specialty Coffee Cupping Form Date Classification: Code 6.00 - Good 8.00-Excellent 9.00-Outstanding Origin 6.25 8.25 9.25 6.50 8.50 9.50 6.75 8.75 9.75 Total Score Sample# Roast Total: Total: Total: Total: Total: Total: Total:

Fragnance/Aroma Flavor Acidity Body Uniformity Clean Cup Overall

Level 6 7 8 9 10 6 7 8 9 10 6 7 8 9 10 6 7 8 9 10 6 7 8 9 10 Dry Crust Intensity Intensity Defects (Subtruct) Total: Total: Quality Aftertaste Total: High High Balance Sweetness Taint=2 # of cups Intensity x 6 7 8 9 10 Low Low 6 7 8 9 10 Fault=4

Final Score Washed Washed Unwashed Unwashed Classification Symbol ClassificationSymbol Classification Symbol Symbol Classification Yirgachefe A WYCA Godere WTPA Yirgachefe A UYCA Harar A UHRA Wenago A WWNA Yeki WYK Wenago A UWNA Harar B UHRB Kochere A WKCA Anderacha WAN Kochere A UKCA Harar C UHRC Gelena Abaya A WGAA Bench Maji WBM Gelena Abaya A UGAA Bale UBL Yirgachefe B WYCB Bebeka WBB Yirgachefe B UYCB Harar E UHRE WenagoWWNB B Kelem Wollega WKM Wenago B UWNB Kelem Wollega UKW Kochere B WKCB East Wollega WEW Kochere B UKCB East Wollega UEW Gelena Abaya B WGABGimbi WGM Gelena Abaya B UGAB Gimbi UGM Sidama A WSDA Kaffa WKF Sidama A USDA Forest A UFRA Sidama B WSDB Sidama B USDB Bench Maji UBM Sidama C WSDC Sidama C USDC Keffa UKF Sidama D WSDD Sidama D USDD Sidama E WSDE Sidama E USDE Limmu A WLMA Jimma A UJMA Limmu B WLMB Jimma B UJMB WGJ Guji UGJ Guji Classification Name Signature Grade Grade Grading Requirements Coordinator: Q Grader 1 : Symbol Preliminary Assessment Grade Cup Quality Q Grader 2 : Speciality 1 Q1 Grade1,Grade2, coffee having > 80.0 Min. 85.0 Speciality 2 Q2 Grade 1 and Grade 2 80.0-74.75

152

Curriculum vitae

Kassaye Tolessa Sherge was born in Shambu, East Wellega, Ethiopia on March 29, 1982. She finished her secondary school at Shambu Senior Secondary School in 2000. She joined Jimma University and graduated with BSc degree in Horticulture in 2004 and recruited in Oromia Agricultural Research Institute, Adami Tulu Research Centre and worked as Junior researcher. In September 2007, she got a Scholarship from NUFFIC and joined Wageningen University, the Netherlands for her MSc study and graduated with MSc degree in Plant Science Specialization in Greenhouse Horticulture in 2009. From 2010 - till present she is serving as lecture at Jimma University College of Agriculture and Veterinary Medicine. In April 2012, She had again a grant from NUFFIC- NICHE ETH-019 and started her PhD research on Biophysical controls of specialty coffee quality in Jimma zone, Ethiopia in the Department of Applied Physical and Analytical Chemistry, Isotope Bioscience Laboratory, Faculty of Bioscience Engineering, Ghent University, under the guidance Prof. dr. ir. Pascal Boeckx.

Teaching and supervision experience (Jimma University)

x Floriculture production and management (2011-2016) x Greenhouse production & management of horticultural crops (2011-2016) x Plant physiology (2010-2013) x Physiology of horticultural crops (2010-2011) x Harvest handling and management of crop plants (2010-2011) x Supervising practical session on lab techniques for PhD students (2015-2016) x Supervising seminar and senior project paper for undergraduate students (2010- 2016).

Laboratory training and working visits

x Distribution (LDPSD) Analysis; Spectral and LDPSD Data Management, and Equipment Maintenance 4th – 7th April 2016, Addis Ababa, Ethiopia x Near infrared spectroscopy (NIRS) - on hands training at RIKILT, Wageningen University (2013) x Training of Trainers (ToT) on Infra-red (IR) Spectroscopy, Laser Diffraction Particle Size

153 Curriculum vitae

x Coffee cooperatives and farmers coffee field visits in Sidama zone (2014) x Coffee quality analysis laboratory: Hawassa and Jimma ECX (2013) x Limu coffee plantation farms and coffee processing station - Jimma zone (2012) x Coffee cooperatives and farmers coffee field visits in Jimma zone (2012) x Oromia coffee cooperative union - Addis Ababa (2012) x Efico coffee company, Antwerpen, Belgium (2012)

Participations on symposium or conferences

x Implication for coffee value chain management: Nuffic closing work shop: Value chain management seminar NICHE ETH-019 on December 2-3, 2015, Adama, Ethiopia

x 25th International Conference on Coffee Sciences September 8-13, 2014 Armenia, Colombia x East African Science workshop on February 14, 2011.

x Jimma University annual conferences (2010)

Supervision of MSc thesis student

x Evaluation of Tomato (Solanum lycopersicum L.) Varieties under Greenhouse and Open Field Condition for Growth, Yield and Quality x Biochemical compositions and cup quality of some CBD resistant Ethiopian coffees (Coffea arabica L.) as influenced by variety and processing methods. x Interactive Effect of Altitude, Shade Level and Harvest Period on Biochemical Composition and Cup Quality of Ethiopian Coffee x Physical and Cup Qualities of Coffee Arabica as affected by Variety and Shade levels at Jimma, Southwest Ethiopia. x Effect of genotype and environment on yield and quality attributes of coffee (Coffea arabica L.) in Manna district, Jimma zone x Influence of Stage and Intensity of Truss Pruning on Fruit Yield and Quality of Tomato (Solanium lycopersicon L.) at Melkassa x Evaluation of Onion varieties for yield and growth in response to plant density in Western Wellega zone, Homa Woreda

154 Curriculum vitae

List of publications

Tolessa, K., Rademaker, M., De Baets, B.,and Boeckx, P. (2016). Prediction of specialty coffee cup quality based on near infrared spectra of green coffee beans. Talanta, 150: 367-374. Tolessa, K., D'heer, J, Duchateau L., and Boeckx, P. (2016). Influence of growing altitude, shade and harvest period on quality and biochemical composition of Ethiopian specialty coffee. Journal of the Science of Food and Agriculture: Doi: 10.1002/jsfa.8114 Yebirzaf, Y., Derbew, B and Tolessa, K. (2016). Tomato (Solanum lycopersicum L.) yield and fruit quality attributes as affected by varieties and growth conditions: Article in press on American-Eurasian Journal of Agricultural& Environmental Sciences. Yebirzaf, Y., Derbew, B and Tolessa, K. (2016). Growth and physiological responses of different tomato (Solanum lycopersicum) varieties in relation to growth conditions. Middle-East Journal of Scientific Research 24 (9): 2904-2908 Tolessa, K., Adugna, D and Getahun, L. (2013). Evaluating seedling establishment of tomato (Lycoperscum esculentum Mill.) varieties as influenced by NaCl stress. Int Curr Res 3:10-14 Adugna, D., Daba G, Bane D., and Tolessa K. (2011). Identification of major causes of postharvest losses among selected fruits in Jimma zone for proffering veritable solutions. Int Curr Res 3:40-43. Girma, K, Tolessa K and Adugna, D. (2011). Evaluating tomato responses: a means of screening varieties to saline soil. Intern J Current Agric Res. 3: 1-11 Geremew, A., Teshome A., Tolessa, K and Amenti C. (2010). Effect of intra- row Spacing on yield of three onion (Allium cepa ) varieties at Adami Tulu Agricultural Research Center (mid rift valley of Ethiopia). Journal of Horticulture and Forestry. 2(1): 007-011. Teshome, A., Amenti, C., Tolessa, K., Fiseha, T. and Geremew, A. (2011). Yield and yield components as influenced by plant density on sweet potato on farmers field, Adami Tulu Jido kombolcha area. American Journal of Experimental Agriculture. 1(2): 007-011 Teshome, A., Tolessa, K. and Amenti, C. (2010). Effect of inter-row spacing with double row arrangement on yield and yield component of tomato (Lycopersicon esculuntum Mill.) at Adami Tulu Agricultural Research Center (Central Rift Valley of Oromia, Ethiopia). African Journal of Agricultural Research. 6(13): 2978-2981.

155