Search
2023 Volume 3
Article Contents
ARTICLE   Open Access    

Progeny selections of coffee cultivar 'Mundo Novo' with potential for the specialty coffee market

  • # These authors contributed equally: Gabriela Ester Ferraz, Marlon Enrique López

More Information
  • Coffee is one of the most popular beverages in the world with a global market worth over ${\$\text{USD}} $ 100 billion. In Brazil, the Arabica coffee (Coffea arabica) cultivar 'Mundo Novo' is one of the most planted cultivars due to its high yield, growth vigor and stability. However, this traditional cultivar is generally not considered a good production source for specialty coffee. Exploring intra-cultivar variation and selecting 'Mundo Novo' progenies with superior sensory quality may provide a viable option for coffee producers targeting the rapidly expanding specialty coffee market. The present study analyzed cup quality attributes, agronomic characteristics and genetic identities of 14 genotypes selected from the putative progenies of 'Mundo Novo' and assessed their potential as specialty coffee in Brazil. Beverage quality was assessed using chemical and sensory analysis according to the methodology proposed by the Specialty Coffee Association (SCA) in six selected genotypes. High sensory quality, with sensory scores above 80 points was observed which qualified them in the very good and excellent quality categories. Genetic identities of the 14 selections were verified using 96 SNP markers, based on 57 reference cultivars from the coffee collection of EPAMIG (Empresa de Pesquisa Agropecuária de Minas Gerais), Brazil. Multivariant and Bayesian clustering analysis generated consistent results and confirmed that all the 14 selections were progenies of 'Mundo Novo'. Among them, the top three genotypes (SGAB 1, SGAB 7 and SGAB 8) were selected as promising genotypes, either for future genetic improvement or for direct use in the production of specialty coffee
  • 加载中
  • Supplemental Table S1 Coffee bean granulometry of 14 coffe genotypes and two controls (Catuaí Vermelho IAC 99 and Mundo Novo)- Year 2015. *Values represented in percentages.
    Supplemental Table S2 Summary of genotypes parameters two by two comparison for the variables height, branch length, number of nodes per branch and stem diameter. The β0’s represent the initial values (intercepts) and the β1’s the increment of the values of the response variables over time (linear regression coefficients).
    Supplemental Table S3 Results of regression analysis of the estimates of  β0’s (intercepts) and β1’s  (linear regression coefficients – slope) for the variables plant height, branch length, number of nodes and stem diameter in 14 genotypes of C. arabica with 95% confidence interval.
    Supplemental Table S4 Result fo SNP genotyping on 14 'Mundo Novo' progenies and 57 reference accessions of Arabica coffee (Coffea arabica),  using nano-fluidic array.
  • [1]

    Conselho dos Exportadores de Café do Brasil (CECAFE). 2021. Exportação anual 2021. www.cecafe.com.br/dados-estatisticos/exportacoes-brasileiras/

    [2]

    Secretaria de Estado de Agricultura, Pecuária e Abastecimento (SEAPA). 2021. Indicadores do café de Minas Gerais, do Brasil e do Mundo, Junho de 2021. www.reformaagraria.mg.gov.br/images/documentos/perfil_caf%C3%A9_junho_2021[1].pdf

    [3]

    Marie L, Abdallah C, Campa C, Courtel P, Bordeaux M et al. 2020. G× E interactions on yield and quality in Coffea arabica: new F1 hybrids outperform American cultivars. Euphytica 216:78

    doi: 10.1007/s10681-020-02608-8

    CrossRef   Google Scholar

    [4]

    Turreira-García N. 2022. Farmers’ perceptions and adoption of Coffea arabica F1 hybrids in Central America. World Development Sustainability 1:100007

    doi: 10.1016/j.wds.2022.100007

    CrossRef   Google Scholar

    [5]

    Jezeer RE, Verweij PA, Boot RG, Junginger M, Santos MJ. 2019. Influence of livelihood assets, experienced shocks and perceived risks on smallholder coffee farming practices in Peru. Journal of Environmental Management 242:496−506

    doi: 10.1016/j.jenvman.2019.04.101

    CrossRef   Google Scholar

    [6]

    Carvalho A, Krug CA, Mendes JET, Antunes Filho H, de Morais H, et al. 1952. Melhoramento do cafeeiro: IV-café Mundo Novo. Bragantia 12:97−130

    doi: 10.1590/S0006-87051952000200001

    CrossRef   Google Scholar

    [7]

    Carvalho CHS (ed). 2008. Cultivares de café: origem, características e recomendações. Brasília: Embrapa. 247 pp

    [8]

    Guerreiro-Filho O, Ramalho MAP, Andrade VT. 2018. Alcides Carvalho and the selection of Catuaí cultivar: interpreting the past and drawing lessons for the future. Crop Breeding and Applied Biotechnology 18:460−66

    doi: 10.1590/1984-70332018v18n4p69

    CrossRef   Google Scholar

    [9]

    Instituto Agronômico de Campinas (IAC). 2021. Cultivares de café desenvolvidas pelo Instituto Agronômico (IAC) e registradas no Ministério da Agricultura, Pecuária e Abastecimento – MAPA (Registro Nacional de Cultivares – RNC). www.iac.sp.gov.br/areasdepesquisa/cafe/tabela_rnc_cultivares_cafe_iac.pdf

    [10]

    Tassone GAT, Nadaleti DHS, Carvalho GR, Pereira FAC, Andrade VT, et al. 2019. Simultaneous selection in coffee progenies of mundo novo by selection indices. Coffee Science 14:83−92

    doi: 10.25186/cs.v14i1.1531

    CrossRef   Google Scholar

    [11]

    Rodrigues CI, Maia R, Miranda M, Ribeirinho M, Nogueira JMF, et al. 2009. Stable isotope analysis for green coffee bean: a possible method for geographic origin discrimination. Journal of Food Composition and Analysis 22:463−71

    doi: 10.1016/j.jfca.2008.06.010

    CrossRef   Google Scholar

    [12]

    de Oliveira Fassio L, Pereira RGFA, Malta MR, Liska GR, Sousa MMM, et al. 2019. Sensory profile of arabica coffee accesses of the germplasm collection of Minas Gerais–Brazil. Coffee Science 14:382−93

    doi: 10.25186/cs.v14i3.1597

    CrossRef   Google Scholar

    [13]

    Sobreira FM, de Oliveira ACB, Pereira AA, Martins AG, Sakiyama NS. 2016. Divergence among arabica coffee genotypes for sensory quality. Australian Journal of Crop Science 10:1442−48

    doi: 10.21475/ajcs.2016.10.10.p7430

    CrossRef   Google Scholar

    [14]

    Borém FM, Figueiredo LP, Ribeiro FC, Taveira JHS, Giomo GS et al. 2016. The relationship between organic acids, sucrose and the quality of specialty coffees. African Journal of Agricultural Research 11:709−717

    doi: 10.5897/AJAR2015.10569

    CrossRef   Google Scholar

    [15]

    Brazil Specialty Coffee Association (BSCA). 2022. http://brazilcoffeenation.com.br/variety/show/id/74

    [16]

    Leroy T, Ribeyre F, Bertrand B, Charmetant P, Dufour M, et al. 2006. Genetics of coffee quality. Brazilian Journal of Plant Physiology 18:229−42

    doi: 10.1590/S1677-04202006000100016

    CrossRef   Google Scholar

    [17]

    Scholz MBS, Figueiredo VRG, Silva JVN, Kitzberger CSG. 2011. Características físico-químicas de grãos verdes e torrados de cultivares de café (Coffea arabica L.) do IAPAR. Coffee Science 6:245−55

    Google Scholar

    [18]

    Bertrand B, Boulanger R, Dussert S, Ribeyre F, Berthiot L, et al. 2012. Climatic factors directly impact the volatile organic compound fingerprint in green Arabica coffee bean as well as coffee beverage quality. Food Chemistry 135:2575−83

    doi: 10.1016/j.foodchem.2012.06.060

    CrossRef   Google Scholar

    [19]

    Malta MR, De Rezende Chagas SJ. 2009. Avaliação de compostos não-voláteis em diferentes cultivares de cafeeiro produzidas na região Sul de Minas Gerais. Acta Scientiarum Agronomy 31:57−61

    doi: 10.4025/actasciagron.v31i1.6629

    CrossRef   Google Scholar

    [20]

    Sunarharum WB, Williams DJ, Smyth HE. 2014. Complexity of coffee flavor: A compositional and sensory perspective. Food Research International 62:315−25

    doi: 10.1016/j.foodres.2014.02.030

    CrossRef   Google Scholar

    [21]

    Tran HTM, Furtado A, Vargas CAC, Smyth H, Lee SL, et al. 2018. SNP in the Coffea arabica genome associated with coffee quality. Tree Genetics & Genomes 14:72

    doi: 10.1007/s11295-018-1282-9

    CrossRef   Google Scholar

    [22]

    Heo J, Adhikari K, Choi KS, Lee J. 2020. Analysis of caffeine, chlorogenic acid, trigonelline, and volatile in crew coffee using high performance liquid chromatography and solid-phase microextraction - gas chromatography-mass spectrometry. Foods 9:1746

    doi: 10.3390/foods9121746

    CrossRef   Google Scholar

    [23]

    International Trade Centre (ITC). 2011. The Coffee Exporter's Guide. 3rd Edition. Geneva: ITC. xvi, 247 pp

    [24]

    Cilas C, Bouharmont P, Boccara M, Eskes AB, Baradat PH. 1998. Prediction of genetic value for coffee production in Coffea arabica from a half-diallel with lines and hybrids. Euphytica 104:49−59

    doi: 10.1023/A:1018635216182

    CrossRef   Google Scholar

    [25]

    Steiger D, Nagai C, Moore P, Morden C, Osgood R, et al. 2002. AFLP analysis of genetic diversity within and among Coffea arabica cultivars. Theoretical and Applied Genetics 105:209−15

    doi: 10.1007/s00122-002-0939-8

    CrossRef   Google Scholar

    [26]

    Tran TMH. 2005. Genetic variation in cultivated coffee (Coffea arabica L.) accessions in northern New South Wales. Thesis. Southern Cross University, Australia

    [27]

    Mishra MK, Tornincasa P, De Nardi B, Asquini, E, Dreos R, et al. 2011. Genome organization in coffee as revealed by EST PCRRFLP, SNPs and SSR analysis. Journal of Crop Science and Biotechnology 14:25

    doi: 10.1007/s12892-010-0035-6

    CrossRef   Google Scholar

    [28]

    Berecha G. 2014. Genetic Diversity, Pollination Ecology and Organoleptic Characteristics of Coffea arabica L., in Ethiopian Moist Forests of Different Management Intensity. Dissertation. KU Leuven, Leuven, Belgium

    [29]

    Gichimu BM. 2016. Selection for Resistance to Coffee Berry Disease, High Yields and Good Quality within Coffea arabica L. Cultivar, Ruiru 11. Thesis. Jomo Kenyatta University of Agriculture and Technology, Kenya

    [30]

    Caetano AL, Barbosa FS. 2019. Probabilidade de ocorrência de chuvas extremas para região de Inconfidentes–MG. Revista Brasileira de Climatologia 25

    doi: 10.5380/abclima.v25i0.62338

    CrossRef   Google Scholar

    [31]

    Souza ACC, Silva ML. 2016. Geoprocessamento aplicado ao levantamento de solos no município de Inconfidentes-MG. Revista Brasileira de Geografia Física 9:200−14

    doi: 10.26848/rbgf.v9.1.p200-214

    CrossRef   Google Scholar

    [32]

    Technical Standards Committee. 2015. SCAA Protocols, Cupping Specialty Coffee. Specialty Coffee Association of America. www.scaa.org/PDF/resources/cupping-protocols.pdf

    [33]

    Lingle TR. 1986. The coffee cupper's handbook: systematic guide to the sensory evaluation of coffee's flavor. Coffee Development Group, Washington

    [34]

    International Organization for Standardization (ISO). 1999. Green coffee – determination of loss mass at 105 ºC. ISO 6673: 2003

    [35]

    Vitorino MD, França AS, Oliveira L, Borges MLA. 2001. Metodologias de obtenção de extrato de café visando a dosagem de compostos não-voláteis. Revista Brasileira de Armazenamento 26:17−24

    Google Scholar

    [36]

    Dishe Z. 1962. General color reaction. In Carbohydrate Chemistry, eds. Whistler RL, Wolfrom ML. New York: Academic Press. pp 477−512

    [37]

    Association Of Official Analytical Chemistry (AOAC). 1990. Official methods of analysis of the Association of Official Analytical Chemists. 15th Edition. Washington

    [38]

    Goldstein JL, Swain T. 1963. Changes in tannins in ripening fruits. Phytochemistry 2:371−83

    doi: 10.1016/S0031-9422(00)84860-8

    CrossRef   Google Scholar

    [39]

    Carvalho VD, Chagas SJR, Chalfoun SM, Botrel N, Juste Junior ESG. 1994. Relação entre a composição físico-química e química do grão beneficiado e a qualidade de bebida do café 1-Atividades de polifenoloxidase e peroxidase, índice de coloração de acidez. Pesquisa Agropecuária Brasileira 29:449−54

    Google Scholar

    [40]

    Loeffler TM, Tekrony DM, Egli DB. 1988. The bulk conductivity test as an indicator of soybean seed quality. Journal of Seed Technology 12:37−53

    Google Scholar

    [41]

    Prete CEC. 1992. Condutividade elétrica do exsudato de grãos de café (Coffea arabica L.) e sua relação com a qualidade da bebida. Doctoral Thesis. Escola Superior de Agricultura Luiz de Queiroz (ESALQ), Piracicaba, Brazil. https://doi.org/10.11606/T.11.1992.tde-20210104-163950

    [42]

    Healey A, Furtado A, Cooper T, Henry RJ. 2014. Protocol: a simple method for extracting next-generation sequencing quality genomic DNA from recalcitrant plant species. Plant Methods 10:21

    doi: 10.1186/1746-4811-10-21

    CrossRef   Google Scholar

    [43]

    Merot-L'Anthoene V, Tournebize R, Darracq O, Rattina V, Lepelley M, et al. 2019. Development and evaluation of a genome-wide Coffee 8.5K SNP array and its application for high-density genetic mapping and for investigating the origin of Coffea arabica L. Plant Biotechnology Journal 17:1418−30

    doi: 10.1111/pbi.13066

    CrossRef   Google Scholar

    [44]

    Zhang D, Vega FE, Solano W, Su F, Infante F, et al. 2021. Selecting a core set of nuclear SNP markers for molecular characterization of Arabica coffee (Coffea arabica L.) genetic resources. Conservation Genetics Resources 13:329−35

    doi: 10.1007/s12686-021-01201-y

    CrossRef   Google Scholar

    [45]

    Wang J, Lin M, Crenshaw A, Hutchinson A, Hicks B, et al. 2009. High-throughput single nucleotide polymorphism genotyping using nanofluidic Dynamic Arrays. BMC Genomics 10:561

    doi: 10.1186/1471-2164-10-561

    CrossRef   Google Scholar

    [46]

    Fluidigm. 2013. Fluidigm® SNP Genotyping User Guide. PN 68000098 Rev I1, South San Francisco, CA, Fluidigm Corporation.

    [47]

    Cruz CD. 2016. Genes Software-extended and integrated with the R, Matlab and Selegen. Acta Scientiarum Agronomy 38:547−52

    doi: 10.4025/actasciagron.v38i3.32629

    CrossRef   Google Scholar

    [48]

    Scott AJ, Knott M. 1974. A cluster analysis method for grouping means in the analysis of variance. Biometrics 30:507−12

    doi: 10.2307/2529204

    CrossRef   Google Scholar

    [49]

    Pritchard JK, Stephens M, Donnelly P. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945−59

    doi: 10.1093/genetics/155.2.945

    CrossRef   Google Scholar

    [50]

    Evanno G, Regnaut S, Goudet J. 2005. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14:2611−20

    doi: 10.1111/j.1365-294X.2005.02553.x

    CrossRef   Google Scholar

    [51]

    Puechmaille SJ. 2016. The program STRUCTURE does not reliably recover the correct population structure when sampling is uneven: subsampling and new estimators alleviate the problem. Molecular Ecology Resources 16:608−27

    doi: 10.1111/1755-0998.12512

    CrossRef   Google Scholar

    [52]

    Li YL, Liu JX. 2018. SᴛʀᴜᴄᴛᴜʀᴇSᴇʟᴇᴄᴛᴏʀ: A web-based software to select and visualize the optimal number of clusters using multiple methods. Molecular Ecology Resources 18:176−77

    doi: 10.1111/1755-0998.12719

    CrossRef   Google Scholar

    [53]

    Jakobsson M, Rosenberg NA. 2007. CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23:1801−6

    doi: 10.1093/bioinformatics/btm233

    CrossRef   Google Scholar

    [54]

    Rosenberg NA, Burke T, Elo K, Feldman MW, Freidlin PJ, et al. 2001. Empirical evaluation of genetic clustering methods using multilocus genotypes from 20 chicken breeds. Genetics 159:699−713

    doi: 10.1093/genetics/159.2.699

    CrossRef   Google Scholar

    [55]

    Dieringer D, Schlötterer C. 2003. Mɪᴄʀᴏsᴀᴛᴇʟʟɪᴛᴇ ᴀɴᴀʟʏsᴇʀ (MSA): a platform independent analysis tool for large microsatellite data sets. Molecular Ecology Notes 3:167−69

    doi: 10.1046/j.1471-8286.2003.00351.x

    CrossRef   Google Scholar

    [56]

    Nadalete DHS. 2020. Banco ativo de germoplasma de Minas Gerais: Avaliação física, sensorial e aceitabilidade do consumidor. Thesis. Universidade Federal de Lavras, Brazil

    [57]

    Allred KF, Yackley KM, Vanamala J, Allred CD. 2009. Trigonelline is a novel phytoestrogen in coffee beans. The Journal of Nutrition 139:1833−38

    doi: 10.3945/jn.109.108001

    CrossRef   Google Scholar

    [58]

    Van Cuong T, Ling LH, Quan GK, Tiep TD, Nan X, et al. 2014. Effect of roasting conditions on several chemical constituents of Vietnam Robusta coffee. The Annals of the University Dunarea de Jos of Galati, Fascicle VI - Food Technology 38:43−56

    Google Scholar

    [59]

    De Maria CAB, Trugo LC, Neto FRA, Moreira RFA, Alviano CS. 1996. Composition of green coffee water-soluble fractions and identification of volatiles formed during roasting. Food Chemistry 55:203−7

    doi: 10.1016/0308-8146(95)00104-2

    CrossRef   Google Scholar

    [60]

    Vaast P, Bertrand B, Perriot JJ, 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 86:197−204

    doi: 10.1002/jsfa.2338

    CrossRef   Google Scholar

    [61]

    Seninde DR, Chambers EIV. 2020. Coffee flavor: A review. Beverages 6:44

    doi: 10.3390/beverages6030044

    CrossRef   Google Scholar

    [62]

    Borém FM. 2008. Pós-colheita do Café. Lavras, Brazil: Universidade Federal de Lavras

    [63]

    Girma B, Gure A, Wedajo F. 2020. Influence of Altitude on Caffeine, 5-caffeoylquinic acid, and nicotinic acid contents of arabica coffee varieties. Journal of Chemistry 2020:3904761

    doi: 10.1155/2020/3904761

    CrossRef   Google Scholar

    [64]

    Ogutu C, Cherono S, Ntini C, Wang L, Han Y. 2022. Comprehensive analysis of quality characteristics in main commercial coffee varieties and wild Arabica in Kenya. Food Chemistry: X 14:100294

    doi: 10.1016/j.fochx.2022.100294

    CrossRef   Google Scholar

    [65]

    Joët T, Laffargue A, Salmona J, Doulbeau S, Descroix F, et al. 2009. Metabolic pathways in tropical dicotyledonous albuminous seeds: Coffea arabica as a case study. New Phytologist 182:146−62

    doi: 10.1111/j.1469-8137.2008.02742.x

    CrossRef   Google Scholar

    [66]

    Cheng B, Smyth HE, Furtado A, Henry RJ. 2020. Slower development of lower canopy beans produces better coffee. Journal of Experimental Botany 71:4201−14

    doi: 10.1093/jxb/eraa151

    CrossRef   Google Scholar

    [67]

    Sant’Ana GC, Pereira LFP, Pot D, Ivamoto ST, Domingues DS, et al. 2018. Genome-wide association study reveals candidate genes influencing lipids and diterpenes contents in Coffea arabica L. Scientific Reports 8:465

    doi: 10.1038/s41598-017-18800-1

    CrossRef   Google Scholar

    [68]

    Malta MR, Pereira RGFA, de Rezende Chagas SJ. 2005. Potassium leaching and eletric conductivity of grain coffee (Coffea arabica L.) exsudate: some factors that may affect these evaluations. Ciência e Agrotecnologia 29:1015−20

    doi: 10.1590/s1413-70542005000500015

    CrossRef   Google Scholar

    [69]

    Saath R. 2010. Qualidade do café natural e despolpado em diferentes condições de secagem e tempos de armazenamento. Thesis. Universidade Estadual Paulista, Brazil

    [70]

    Mazzafera P, Gonçalves KV, Shimizu MM. 2002. Extraction and activity determination of polyphenoloxidase in coffee. Scientia Agrícola 59:695−700

    doi: 10.1590/s0103-90162002000400012

    CrossRef   Google Scholar

    [71]

    Antunes Filho H, Carvalho A. 1954. Genética de Coffea: XIX-Mutação somática afetando a cor das folhas em café. Bragantia 13:XI−XII

    doi: 10.1590/S0006-87051954000100029

    CrossRef   Google Scholar

    [72]

    Krug CA. 1949. Mutações em Coffea arabica L. Bragantia 9:1−10

    doi: 10.1590/S0006-87051949000100001

    CrossRef   Google Scholar

    [73]

    Lashermes P, Combes MC, Robert J, Trouslot P, D'Hont A, et al. 1999. Molecular characterisation and origin of the Coffea arabica L. genome. Molecular and General Genetics MGG 261:259−66

    doi: 10.1007/s004380050965

    CrossRef   Google Scholar

  • Cite this article

    Ferraz GE, Lópes ME, Gonçalves FMA, Malta MR, Lima RRD, et al. 2023. Progeny selections of coffee cultivar 'Mundo Novo' with potential for the specialty coffee market. Beverage Plant Research 3:1 doi: 10.48130/BPR-2023-0001
    Ferraz GE, Lópes ME, Gonçalves FMA, Malta MR, Lima RRD, et al. 2023. Progeny selections of coffee cultivar 'Mundo Novo' with potential for the specialty coffee market. Beverage Plant Research 3:1 doi: 10.48130/BPR-2023-0001

Figures(4)  /  Tables(2)

Article Metrics

Article views(5576) PDF downloads(682)

ARTICLE   Open Access    

Progeny selections of coffee cultivar 'Mundo Novo' with potential for the specialty coffee market

Beverage Plant Research  3 Article number: 1  (2023)  |  Cite this article

Abstract: Coffee is one of the most popular beverages in the world with a global market worth over ${\$\text{USD}} $ 100 billion. In Brazil, the Arabica coffee (Coffea arabica) cultivar 'Mundo Novo' is one of the most planted cultivars due to its high yield, growth vigor and stability. However, this traditional cultivar is generally not considered a good production source for specialty coffee. Exploring intra-cultivar variation and selecting 'Mundo Novo' progenies with superior sensory quality may provide a viable option for coffee producers targeting the rapidly expanding specialty coffee market. The present study analyzed cup quality attributes, agronomic characteristics and genetic identities of 14 genotypes selected from the putative progenies of 'Mundo Novo' and assessed their potential as specialty coffee in Brazil. Beverage quality was assessed using chemical and sensory analysis according to the methodology proposed by the Specialty Coffee Association (SCA) in six selected genotypes. High sensory quality, with sensory scores above 80 points was observed which qualified them in the very good and excellent quality categories. Genetic identities of the 14 selections were verified using 96 SNP markers, based on 57 reference cultivars from the coffee collection of EPAMIG (Empresa de Pesquisa Agropecuária de Minas Gerais), Brazil. Multivariant and Bayesian clustering analysis generated consistent results and confirmed that all the 14 selections were progenies of 'Mundo Novo'. Among them, the top three genotypes (SGAB 1, SGAB 7 and SGAB 8) were selected as promising genotypes, either for future genetic improvement or for direct use in the production of specialty coffee

    • Brazil is the world's leading producer and exporter of coffee and the coffee industry contributes a total of US${\$} $ 6.2 billion to Brazil's economy[1]. Minas Gerais is one of the major coffee producing regions in Brazil. With a territorial area of 582,586 km2, Minas Gerais produced 1,400,640 tons of coffee, accounting for 47.8% of the total production in Brazil[2]. These heterogenous environments enable the state of Minas Gerais to produce quality coffees with a greater diversity of flavor and aroma. Furthermore, the heterogenous production system in Minas Gerais requires a greater diversity of coffee cultivars adapted to different soil characteristics, nutrient availability, climate conditions and cropping systems.

      For any new coffee plantations, the selection of cultivars most suitable for the local environment and cropping system needs careful consideration and planning. This is because coffee is a perennial crop with a lifespan of 15 to 25 years and a large investment is necessary to establish a new plantation[3]. In addition, the introduction of new cultivars often requires changes to crop management practice which is usually associated with additional labor or equipment and the input of additional resources. In other cases, farmers may also be culturally attached to traditional varieties and thus are hesitant to accept cultivars that they do not know[4]. These factors influence the adoption of new cultivars by coffee farmers who may prefer to use known cultivars that have a predictable performance[4,5].

      The coffee cultivar 'Mundo Novo' was discovered in the 1940s in Brazil when the demand for cultivars was focused on high productivity[6]. Phenotypically, 'Mundo Novo' has an excellent vegetative vigor, large number of branches, and a well-developed root system, which contributes to high yields and adaptability[7,8]. These favorable characteristics made 'Mundo Novo' one of the most popular cultivars widely grown by coffee producers in Brazil, and the cultivar remains popular today[9,10].

      However, the demand for specialty coffee is rapidly increasing across domestic and international markets where consumers seek exceptional taste and aroma as well as balanced characteristics of sweetness, acidity and body[11]. Productivity is no longer the single dominant factor determining cultivar adoption by coffee producers. The sensory quality of 'Mundo Novo' marginally fits into the category of gourmet coffee[12,13]. In contrast, cultivars such as 'Maragogipe'[7], 'Bourbon'[14], 'Pacamara'[12] and 'Geisha'[15], have been recognized in the specialty coffee market because of their superior cup quality.

      Breeding new cultivars for dual high-yield, higher-cup quality is technically challenging, because there are complicated factors involved in cup quality[16]. Both genetic[17] and environmental[18] factors affect the plant's physiological processes, which determine the formation of a broad array of chemical compounds in different combinations and amounts[19]. Moreover, good cup quality is a result of proper post-harvest processing and roasting, which affects the formation of volatile and non-volatile compounds that are important for coffee aroma and flavor[20,21].

      Non-volatile organic compounds are produced in multiple phases of coffee production, from green (raw) beans to the brewed coffee we consume. Studies have shown that some non-volatile compounds make up the taste or flavors[22].

      The grading and classification of coffee in the global market are based on multiple criteria: the roast appearance and cup quality (flavor, flagrance, acidity, and body), bean physiognomies (size, shape and color), the density of beans and number of defects[23]. This complexity, with the addition of other agronomic traits, such as yield, disease resistance, and stability, make it difficult for breeders to meet multiple breeding objectives simultaneously.

      On the other hand, a significant amount of intra-cultivar variation exists in Arabica coffee, even though C. arabica is a predominantly reproduced via self-pollination[2427]. This intra-cultivar variation was mainly due to the accumulation of mutations and residue heterozygosity[24,25]. In addition, outcrossing is commonly found in Arabica coffee, which also contribute to intra-cultivar variability in some cultivars[28]. These intra-cultivar variations have been explored by coffee breeders to select new cultivars with high yields and improved disease resistances. For example, Gichimu[29] reported the existence of a significant yield variation among 'Ruiru 11' siblings and practiced intra-cultivar selection for high yields. Tassone et al.[10] took advantage of the intra-cultivar variation in 'Mundo Novo' and selected promising progenies with high yield capacity, bean size, plant height, upper and lower canopy diameter, stem diameter and plant vigor. Nonetheless, so far, few studies have been carried out on the utilization of intra-cultivar heterogeneity to improve cup quality.

      The present study aims to select genotypes with a superior cup quality from the progenies of a traditional cultivar 'Mundo Novo', to develop a new germplasm with specialty coffee quality and wide adoption. A combined approach of sensory testing, biochemical analysis and SNP genotyping was performed to assess cup quality, chemical composition and verify the genetic background of the putative progenies of 'Mundo Novo'. The identified selections have a superior beverage quality, demonstrating their potential for direct adoption for breeding cultivars with improved cup quality. This information provides a good example of the rapid identification of promising coffee genotypes by the exploration of intra-cultivar genetic diversity in Arabica coffee.

    • To verify the genetic potential for cup quality, 14 seven-year-old coffee genotypes were used as single plants, derived from a single mother tree of 'Mundo Novo' at the Sítio Gabriela. The experimental site is in the municipality of Inconfidentes, located in southwestern region of Minas Gerais, Brazil. The average annual rainfall in the municipality is 1,471.6 mm, with wet summer and dry winter seasons[30]. The location is in a section of the 'Serra da Mantiqueira' that is part of the Atlantic Forest Biome[31]. The altitude of the property is 1007 (m.a.s.l) and the soil is classified as a sandy clay.

      The genotypes were selected by mass selection within a stand of the 'Mundo Novo IAC 376/4' cultivar based on their phenotypic variations among the progenies. The selected plants stood out in terms of vigor, bud color and large red fruits with a high percentage of grains retained in sieves equal to or greater than 16 (in an evaluation carried out in 2015, where the progenies obtained an average of 91.4% and the control cultivar 'Mundo Novo' obtained 81% - Supplemental Table S1). The original experiment was installed in December 2009 with a spacing of 1.5 m between plants and 2.5 m between rows. The implementation and management were carried out in accordance with regional technical recommendations for the coffee crop.

    • For the phenotypic characterizations, the vegetative growth of the single 14 genotypes was evaluated over 12 consecutive months (from September 2018 to August 2019). For that, four branches of the middle third of each genotype were selected and particular characteristics of the aerial part of the plant were chosen. The morphological characteristics evaluated were the stem diameter (cm), the plant height (cm), the length of plagiotropic branches (cm) and the number of nodes for each plagiotropic branch.

    • Of the total 14 genotypes evaluated, only six single plants had enough fruit production to carry out the sensory analyses. The fruits were collected in the cherry stage, by selective harvesting, followed by drying by natural processes (described below). Two additional cultivars, 'Catuaí Vermelho IAC 99' and 'Mundo Novo IAC 376/4', from the same property were included in the analysis as controls. The fruits were harvested in 2019 and, within six hours after beginning the processing, they were taken to a concrete yard and evenly distributed on sieves (with wooden frames and wire mesh with 2.00 mm × 1.00 mm diameter openings), while maintaining separate samples. The samples were considered dried when the grains reached a moisture content of 11% to 12%. Then, the samples of dried coffee cherries, from natural processes were further processed (dehusked) and sent for laboratory analysis.

      Sensory analysis was performed by three accredited tasters, according to the methodology proposed by the Specialty Coffee Association (SCA)[32], corresponding to three repetitions. The roasting process complied with the Agtron 'Gourmet' scale (63). According to this methodology, each attribute evaluated (fragrance/aroma, flavor, aftertaste, acidity, body, balance, sweetness, uniformity, cleanliness and scoring) receives grades from 0 to 10, according to the intensity in the samples based on the terminology presented by Lingle[33]. According to the SCA classification, coffee samples that achieve a score of less than 80 points do not qualify as special, 80 to 84.99 are classified as very good, 85 to 89.99 as excellent and above 90 points is considered exceptional.

    • The physical-chemical analysis was carried out at the 'Empresa de Pesquisa Agropecuaria de Minas Gerais' (EPAMIG) Laboratory 'Dr. Alcides Carvalho', in the municipality of Lavras, MG. From each of the collected samples, four replicates were performed, characterizing laboratory quadruplicates. Moisture content was determined by the oven method at 105 ± 1 °C for 16 ± 0.5 h, according to the international standard method of ISO 6673[34]. The grain samples from the plants were ground for about one minute in a mill model TE 631/2, brand Tecnal. After grinding, they were placed in plastic containers and stored in a freezer at a temperature of −18 °C, until the analysis was performed.

      The non-volatile compounds such as trigonelline, chlorogenic acid, 5-caffeoylquimic acid (5-CQA) and caffeine were determined by High Performance Liquid Chromatography (HPLC), according to a methodology adapted from Vitorino et al.[35] by Malta & Chagas[19]. Samples of ground roasted coffee (0.5 g) were extracted with 100 mL of boiling distilled water, stirring for 3 min in a water bath at 100 °C. The system is coupled to a Shimadzu UV/visible spectrophotometric detector (model SPD-10A) connected by an interface (CBM-101) to a microcomputer for data processing. The analysis conditions used were 1 mL·min−1 flow; mobile phase: methanol, water, and acetic acid (20:80:1); room temperature; wavelength 272 nm. The concentration of compounds was determined by the relationship between the peak areas of caffeine, trigonelline, and chlorogenic acid in the sample and their respective Sigma standards known concentrations.

      Total sugars were determined by the Antrone method[36]. Lipid content was quantified by extraction with ethyl ether in a Soxhlet apparatus[37]. Polyphenols were hotly extracted by the method of Goldstein & Swain[38]. Protein determination was performed using the method of AOAC. Polyphenol oxidase enzymatic activity (U·min−1·g−1 of sample) was detected using the method adapted by Carvalho et al.[39]. The electrical conductivity was determined by the method proposed by Loeffler et al.[40]. Potassium ion leaching was determined according to the method of Prete[41].

    • For verification of genetic identity using SNP genotyping, fully expanded leaves at the third pair were collected from plagiotropic branches from the middle third of coffee plants. The leaves were placed in 50 ml Falcon-type polypropylene tubes, with samples individually separated for each of the 14 genotypes. The same process was carried out for 62 accessions collected in the active germplasm bank (BAG) of the EPAMIG. These accessions were included to serve as candidate parents and reference cultivars for the 14 SGAB genotypes.

      For DNA extraction, the coffee leaves were pulverized, using liquid nitrogen, and soon after, 200 mg of plant material was stored in 2 ml 'Eppendorf' tubes. The DNA extractions were performed according to the protocol CTAB described by Healey et al.[42], with this change in the volume of plant material. Quantity and quality analysis were performed using a Nanovue® spectrophotometer (NanoVue GE Healthcare). 1.5 μL of each sample were used, observing the DNA concentration and the purity ratios at A260/280 and A260/230, prioritizing values between 1.8 to 2.0 for the two ratios. Subsequently, the samples were sent to the Laboratory of Sustainable Perennial Crops, United States Department of Agriculture, Agricultural Research Service (USDA – USA).

      A core set of 96 SNP markers was used for genotyping. These SNPs were selected from the published SNP sequences by Merot-L'anthoene et al.[43]. The detailed information, including criteria of validation, selected SNPs and their flanking sequences, were presented in Zhang et al.[44]. The Fluidigm nanofluidic genotyping system (South San Francisco, CA, USA) was used to evaluate the putative SNP markers for genotype identification. The Assay Design Group at Fluidigm Corp. (South San Francisco, CA, USA) designed and manufactured the putative SNP primers for competitive allele-specific PCR, enabling bi-allelic scoring of SNPs at specific loci (KBioscience Ltd, Hoddesdon, UK).

      SNP genotyping was performed using the Fluidigm 96.96 Dynamic ArrayTM (Fluidigm, San Francisco, CA, USA). Each 96.96 Dynamic Array can run 96 samples against 96 SNP assays generating a total of 9.216 data points in a single experiment. One key feature of this protocol is the inclusion of a specific targeted amplification (STA) reaction[45], which allows the enrichment of template molecules for each individual Integrated Fluidic Circuit® (IFC) reaction that facilitates the multiplexing during genotyping. An advantage to STA is that it allows the use of limited or low-quality DNA samples and reduces bias that may occur when samples are loaded to the 96 sample wells of the IFC. Because coffee leaf tissues contain high levels of polysaccharides and polyphenolic compounds that can potentially inhibit PCR amplification, the STA step is strongly recommended. The STA reaction was performed as described in the Fluidigm SNP Genotyping User Guide, PN 68000098 Rev I1[46]. The STA master mix consisted of 2.5 µL of TaqMan® Taq polymerase (Life Technologies, Carlsbad, CA, USA), PreAmp Master Mix (2×), 1.25 µL of Pooled assay mix (0.2×), and 1.25 µL of genomic DNA for a total reaction volume of 5.0 µL. PCR was performed with an initial denaturation step of 95 °C for 10 min, followed by 14 cycles of a 2-step amplification profile consisting of 14 sec at 95 °C and 4 min at 60 °C. The resulting amplified DNA was then diluted 1:5 in TE buffer in order to reduce the concentration of any remaining PCR by-products. Samples were then genotyped using the nanofluidic 96.96 Dynamic ArrayTM IFC (Integrated Fluidic Circuit; Fluidigm Corp.). The 96.96 Dynamic Array IFC for SNP genotyping was described by Wang et al.[45] End-point fluorescent images of the 96.96 IFC were acquired on an EP1TM imager (Fluidigm Corp.). The data was recorded with Fluidigm Genotyping Analysis Software (Fluidigm, San Francisco, CA, USA).

    • The analysis of stem diameter, plant height, length of plagiotropic branches and number of nodes for each plagiotropic branch, 1st degree linear regression equations were adjusted for each plant and confidence intervals for the parameters of the equation, intercept and regression coefficient were obtained. The parameters of the equations were compared, to verify if the intercept and the regression coefficient were different for the evaluated plants. Comparisons were made between plants, two by two. For each pair of plants, the following regression models were used.

      Plant 1: Y1i = β01 + β11 Xi + e1i

      where: Y1i is the observed value of plant 1 in month i, with i = 1, 2, …, 12; β01 is the intercept of the equation for plant 1, which represents the value of the response variable in month 0, that is, when the straight line passes through the y axis. This value represents the initial value of the response variable; β11 is the regression coefficient, which represents the slope of the line. It can be interpreted as the growth rate of the response variable, that is, the higher its estimated value, the greater the increase in the response variable over time; Xi refers to the month of assessment, where Xi = 1, 2, …, 12, for i = 1, 2, …, 12; respectively; e1i is the random error associated with each observation Y1i, where e1i ~N(0,σ2 ).

      Plant 2: Y2i = β02 + β12 Xi + e2i

      where: Y2i is the observed value of plant 2 in month i, with i = 1, 2, …, 12; β02 is the intercept of the equation for plant 2, which represents the value of the response variable in month 0, that is, when the straight line passes through the y axis. This value represents the initial value of the response variable; β12 is the regression coefficient, which represents the slope of the line. It can be interpreted as the growth rate of the response variable, that is, the higher its estimated value, the greater the increase in the response variable over time; Xi refers to the month of evaluation, with, for, respectively; e2i is the random error associated with each observation Y2i, where e2i ~N(0,σ2).

      After adjusting the equations, the estimated values ​​and confidence intervals (CI) obtained for each parameter of the equation were used to compare the parameters of the equations two by two: (i) for each pair of plants, the intercepts were compared, that is, the initial values ​​of the plant response variables. Thus, CI were used to check whether β01 = β02. The results were summarized in a table, using equal letters for equal parameters. Plants with higher β0's are those with higher initial values ​​of the response variables; (ii) for each pair of plants, the regression coefficients were also compared, that is, the rates of increase ('growth') in the variable responses of the plants over time. Thus, CI were used to check whether β11 = β12. The results were summarized in a table, using equal letters for equal parameters. Plants with higher β1's are those that showed higher increases or 'growth' rate in the values ​​of the response variables over time.

    • Univariate analyses for chemical and sensory data were performed with the GENES software[47] and, if significant differences were detected by the analysis of variance, the Scott-Knott[48] test was used to compare the means.

      For the chemical variables, principal component analysis was performed, and Pearson's correlation coefficients were calculated to evaluate the relationship among these variables.

    • Population structure of the coffee samples was determined using a model-based Bayesian cluster analysis software STRUCTURE v2.3.4[49]. Admixture model was used, and the analysis was carried out without assuming any prior information about the genetic groups or geographic origins of the samples. Ten independent runs were assessed for each fixed number of clusters (K value) ranging from 1 to 10, with a burn-in length of 50,000 followed by 100,000 iterations. The most probable number of clusters was detected using the methods of Evanno et al.[50] as well as the method of Puechmaille[51]. The computation was performed using the on-line program STRUCTURE SELECTOR[52]. For the Puechmaille method, the mean membership threshold value was set at 0.5. The software CLUMPP 1.1[53] was used to find optimal alignments of independent runs and the output was used directly as input into a program for cluster visualization DISTRUCT 1.1[54].

      To illustrate the genetic relationships among germplasm groups and sub-groups, distance-based multivariate analysis was performed on the individual data. A cluster analysis using the UPGMA (Unweighted Pair Group Method with Arithmetic Mean) was performed to further examine the genetic relationship among the SGAB genotypes with other reference cultivars, including 'Mundo Novo'. Pairwise genetic distance based on proportion of shared alleles was determined following the program MICROSATELLITE ANALYZER[55]. UPGMA dendrogram was generated from the resulting matrix of Nei's distance using the computer program PHYLIP version 3.697 (http://evolution.genetics.washington.edu/phylip.html). The dendrogram was visualized with the program FigTree v1.4.3 (http://tree.bio.ed.ac.uk/software/figtree/).

    • The result of sensory characterization is shown in Fig. 1. Only six out of the 14 genotypes produced sufficient cherries for sensory analyses. All six tested SGAB genotypes were classified as specialty coffee following SCA protocol[32]. Among them, SGAB 7 was ranked as the best genotype, with a sensory score of 87.50, thus was rated as excellent cup quality. All six genotypes, except SGAB 13, had a sensory performance score significantly higher than the controls ('Catuaí' and 'Mundo Novo'), thus were classified as excellent or exceptional according to the SCA's protocol. Description of the sensory profile is presented in Table 1, which was based on the nuances identified by the Q-graders (Tasters) in an evaluation carried out by the cup tasting. These genotypes presented floral and fruity notes characteristics that present high scores and categorized as excellent or exceptional quality. Several other nuances in different intensities were also identified.

      Figure 1. 

      Total score according to the SCA methodology, obtained in univariate sensory analysis for six genotypes and two control cultivars ('Catuaí Vermelho IAC 99' and 'Mundo Novo IAC 376/4' - MN) referring to the 2018/2019 crop year. Means followed by the same letter do not differ by the Scott Knott test (1974) (P < 0.05).

      Table 1.  Nuances identified by three tasters based on the methodology described by SCA using the six coffee genotypes provided and two control cultivars.

      GenotypesSensory description
      SGAB 7Vivid acidity, creamy and soft body, floral, honey, Swiss lemonade, fruity, long and sweet finish.
      SGAB 8Vivid and sweet acidity, creamy and soft body, yellow fruits, floral, honey, brown sugar, long and sweet finish.
      SGAB 1Bright and lively acidity, creamy and velvety body, floral flavor, yellow fruits, honey, molasses, brown, long and pleasant finish.
      SGAB 10Sweet acidity, creamy body, brown, caramel, long finish.
      SGAB 9Vivid acidity, creamy body, high sweetness, honey, caramel, milk chocolate, long and pleasant finish.
      Mundo Novo IAC 376/4Sweet acidity, creamy body, high sweetness, yellow fruit, caramel, brown, milk chocolate, long finish.
      SGAB13Medium acidity, creamy body, caramel, milk chocolate, brown, long and sweet finish.
      Catuaí Vermelho IAC 99Medium acidity, creamy body, clean, caramel, milk chocolate, brown, long finish.
      Source: adapted from Nadaleti[56].

      In cup tastings, a score higher than 85 requires a greater harmonization of the qualities presented by the coffee (bean quality, chemical composition, roasting processes, grinding and others). Therefore, it is very difficult to reach the high score and be categorized as excellent and exceptional. Consequently, coffees that reach scores above 90 stand out and receive the highest market prices[16]. In the present study, we observed significant intra-cultivar variation of cup quality among the six tested SGAB genotypes, which are putative progenies of 'Mundo Novo', a popular traditional variety in Brazil. Among them, the genotypes SGAB 1, SGAB 7 and SGAB 8 qualify as excellent with unique attributes, such as floral and fruity determinants, which contributed to their superior cup quality.

    • Significant phenotypic variation was observed among the 14 SGAB genotypes evaluated from the seven years old trees. These traits included: length of the plagiotropic branches; number of nodes in the plagiotropic branches; stem diameter and plant height (Supplemental Table S2). The length ranged from 0.39 to 0.77 m. The number of nodes range from 12.96 to 27.64. The same occurred with stem diameter and height, with variations from 0.26 to 0.57, and 2.76 to 3.65 m, respectively. Likewise, the increments showed variation and changes in genotype rankings (Supplemental Table S2). As for the length of plagiotropic branch, the SGAB 7 genotype differed significantly from the others (Fig. 2a), with an average increase rate of 0.006. The genotype SGAB 13 obtained increment values among the smallest both for stem diameter, 0.001, and height, 0.008 (Fig. 2c). Lower stem and height growth may be a consequence of the plant's difficulty in taking advantage of nutrients, as well as a possible negative interaction with environmental factors in the past cropping years. Growth values among the highest for all variables were observed in relation to genotype SGAB 4. The highlight was for the variable height with 0.036 increment, showing the highest value (Fig. 2d). This fact may represent a greater plasticity and greater capacity for homeostasis of this material, which obtained performance within the highest growth values for the observation period.

      Figure 2. 

      Confidence interval (95%) of β1's estimates (linear regression coefficients - slope) for the variables (a) branch length, (b) number of nodes, (c) stem diameter and (d) plant height in 14 genotypes of C. arabica. Confidence interval values that are not overlapped indicate that estimates differ statistically from each other (Supplemental Table S3).

    • Chemical analyzes were performed for six genotypes and two control cultivars ('Catuaí Vermelho IAC 99' and 'Mundo Novo IAC 376/4'). The contents of trigonelline, 5-CQA chlorogenic acid, caffeine, total sugars, lipids, polyphenols, and proteins were quantified, as well as the enzymatic activity of polyphenol oxidase, electrical conductivity and potassium leaching. Based on analysis of variance and subsequent mean comparison, using Scott-Knott test, significant difference was found among the genotypes for nine out of the ten variables (Table 2).

      Table 2.  Chemical characterization of six genotypes and two control cultivars ('Catuaí Vermelho IAC 99' and 'Mundo Novo IAC 376/4') related to the 2018/2019 crop year.

      GenotypeECLKTSPPOPOLPLTrigCAC
      SGAB734.20 d28.49 f7.96 d46.50 a6.38 b11.68 b12.91 a1.13 b3.91 c1.16 c
      SGAB841.09 b46.02 b8.51 c44.08 a6.31 b12.29 a10.62 b1.15 b3.84 c1.17 c
      SGAB137.40 c30.71 e8.90 b45.30 a7.10 a11.65 b12.79 a1.17 b3.69 d1.07 d
      SGAB1039.68 b43.06 c9.40 a44.43 a6.15 b12.56 a10.61 b1.09 b3.74 d1.25 b
      SGAB955.02 a56.87 a9.28 a45.41 a5.85 b12.53 a10.86 b1.15 b3.74 d1.34 a
      Mundo Novo 376/443.08 b36.17 d8.55 c45.54 a7.32 a12.24 a10.44 b1.42 a5.23 a1.36 a
      SGAB1341.6 b42.76 c8.69 c42.44 a6.02 b12.20 a12.18 a1.06 b3.91 c1.27 b
      Catuaí Vermelho IAC 9932.18 e25.73 g8.30 c47.92 a5.70 b12.16 a10.64 b1.12 b4.27 b1.11 d
      EC, Electrical Conductivity; LK, Potassium Leaching; TS, Total Sugars; PPO, Polyphenol oxydase; POL, Polyphenols, P, Proteins; L, Lipids; Trig, Trigonelin; CA, Chlorogenic Acid; C, Caffeine. Different letters represent statistical significance according to Scott-Knott test.

      Regarding trigonelline content, the cultivar 'Mundo Novo IAC 376/4' had the highest value of 1.42 which is significantly different from the other accessions. However, 'Mundo Novo IAC 376/4' had the lowest score of cup quality (83.7−84.2). In contrast, the genotype SGAB7 had a trigonelline content of 1.13, received the highest score for cup quality (87.5). Trigonelline is the second most important alkaloid next to caffeine and it is one of the main flavor precursors in coffee[5759]. The present result suggests that in Arabica coffee genotypes, the trigonelline content may not have a simple relationship with sensory quality. In addition, it is possible that there is a significant genotype x environment interaction for trigonelline content in the arabica coffee used in this experiment. Relationship between the two variables, as revealed by the present study, could be site specific.

      Cultivar 'Mundo Novo IAC 376/4' also had the highest content of 5-CQA chlorogenic acid of 5.23%, which is significantly higher than all the six SGAB genotypes (Table 2). This result is compatible with the result of cup quality testing. During roasting, chlorogenic acids decompose to form caffeic and quinic acids, which are associated with astringency, bitterness and acidity in beverage[60]. Therefore, chlorogenic acids are strongly associated with low quality coffee, despite it is beneficial effect to human health as an antioxidant. Low chlorogenic acids content is one of the breeding objectives for improving the quality of Arabica coffee[61].

      Caffeine content varies significantly among the tested genotypes, ranging from 1.07% to 1.36%, with Cultivar 'Mundo Novo IAC 376/4' having the highest caffeine content. This range of variation is consistent with previous reports[6264]. The lower levels of caffeine content is a common goal for Arabica coffee breeding, because a high level of caffeine content is generally associated with inferior coffee quality.

      The relationship among the chemical compounds was analyzed using principal component analysis and the biplot is presented in Fig. 3a. The first two principal components explained 67.03% of the total variation of chemical parameters and of the sensory score (41.08% for PC1 and 25.95% for PC2). The variables that contributed most to PC1 were proteins and potassium leaching, whereas for PC2, the variables 5-CQA chlorogenic acid and trigonelline were main contributors.

      Figure 3. 

      (a) Principal components and (b) Pearson correlation analysis of chemical variables of six coffee genotypes evaluated. Polyphenol oxydase (ppo); Total Sugars (t_s); Potassium Leaching (l_k), Electrical Conductivity (e_c).

      The data obtained were also subjected to estimates of Pearson's correlation coefficient (Fig. 3b). High correlations were found between the variables electrical conductivity and potassium leaching with proteins and caffeine, 0.60; 0.65 and 0.74; 0.65, respectively (Fig. 3b) as well as proximity between the biplot vectors for the same variables (Fig. 3a). The variable total sugars also had a high magnitude correlation with potassium leaching (0.67), electrical conductivity and proteins (both with 0.62). These compounds represented determinant variables in the composition of PC1. Low values of these compounds were observed in the genotypes with higher sensorial quality. This finding, in view of the above context, allows us to infer that this main component consists of variables that, at high rates, affect the cup quality.

      In the correlation analysis (Fig. 3b), high correlation was observed among trigonelline, chlorogenic acid (5-CQA) with polyphenols (0.86 and 0.76, respectively), which is consistent with the proximity observed in their respective biplot vectors. Trigonelline and chlorogenic acids were the variables that most contributed to the formation of PC2, which were related to genotypes with lower sensory quality. Some key non-volatile compounds include alkaloids (caffeine and trigonelline) and chlorogenic acid (CGA). Both caffeine and CGA contribute to the bitter flavor, but CGA, which degrades rapidly and forms phenolic compounds, also produces astringent and acidic flavors[22].

      Sugars are related to higher sensory beverage quality, especially sucrose, however this compound is also the largest carbohydrate transported in coffee phloem[65] being a source of energy and carbon skeletons for other structural molecules[66]. Genotype SGAB7 that obtained the highest score in sensory quality was the one that obtained the lowest levels of total sugars, which supports the hypothesis that sugars, instead of being accumulated, were directed to the synthesis of other molecules, which may be involved as the growth and development of the plant, and also in the cup quality.

      As for the levels of lipids, a significant difference was found among the genotypes. The highest values ​​were observed in genotypes SGAB 1, SGAB 7 and SGAB 13. Lipids are recognized as quality indicators[67] and for this reason, this variable presented a correlation of 0.54 (Fig. 3b) with the final sensory score and close proximity – which almost overlaps – of their respective vectors in the biplot (Fig. 3a). This is due to the lipid protection of the internal structures of the cells, which contributes to less volatilization of aromas during the roasting process.

      The results found for the lipid component in genotypes SGAB 1 and SGAB 7 corroborate with lower index of electrical conductivity and potassium leaching, because the smaller the damage caused to cell membranes, the less solutes are leached and, consequently, the indices of electrical conductivity and leaching potassium will be smaller[68]. Therefore, the high positive correlation between these two variables was found (0.91), as well as the proximity of their respective vectors in the biplot (Fig. 3a & b).

      The protein content varied from 11.65% till 12.56%. High protein contents have been associated with lower beverage quality[69]. In genotypes SGAB 7 and SGAB 1 that had the highest sensory scores, lowest protein content was found, differing from genotypes SGAB 9 and SGAB 10 that had high protein values ​​and fit into the groups that had the lowest scores. Consistent with these results is the high negative correlation of −0.83 between the protein and lipid variables (Fig. 3b), as both have already been reported in the literature as indicators of opposite beverage quality.

      The enzymatic activity of polyphenol oxidase (PPO) values ranged from 42.44 to 47.92 U·min−1·g−1 of the sample. Enzyme activity below 55.99 U·min−1·g−1 of the sample was associated with the lowest quality of the beverage[39], which was not observed in this study. According to the results, the evaluated genotypes did not differ significantly from the 'Mundo Novo IAC 376/4' cultivar in relation to PPO activity. The enzymatic activity of PPO had a negative correlation (−0.58) with the variable potassium leaching (Fig. 3b). The lower activities of this enzyme are related to the degradation of cell membranes due to the generation of quinones from the oxidation of phenols that react with this enzyme, decreasing its concentration[70]. In the situation of compromised cell structure, high levels of potassium leaching are observed, which justifies the negative correlation found.

      The polyphenol contents were observed from 5.70% to 7.32%. Lower levels of polyphenols are associated with better maturation of coffee grain, consequently, better sensory quality. However, it was not the genotypes that reached the highest sensory scores that necessarily obtained the lowest concentrations of phenolic compounds, as in genotype SGAB 9 and 'Catuaí Vermelho IAC 99', because it was performed the selective harvesting in which all the fruits collected were mature.

    • Based on the genotyping result of 96 SNP markers (Supplemental Table S4), the UPGMA dendrogram (Fig. 4a) depicted the relationship among the 14 SGAB genotypes and the 57 reference cultivars. Despite the segregation of morphological characteristics among the 14 genotypes, they all formed a tight sub-cluster based on SNP profiles, suggesting that they have a common parentage. The cultivar 'Mundo Novo 379/19' was closely grouped with the 14 SGAB genotypes, demonstrating the high similarity between parent and progenies. However, even the progenies supposed to be derived through self-fertilization, there is clear intra-cultivar variation as shown by the separated into small clusters (Fig. 4b).

      Figure 4. 

      (a) Genetic diversity and (b) structure population of SGAB coffee genotypes compared with EPAMIG coffee germplasm.

      Another accession that closely grouped with the 14 SGAB genotypes is 'Purpuracea' (Fig 4a, b). 'Purpuracea' is a natural mutation of Maracaturra, which is the result of a natural cross of 'Maragogype' and 'Caturra'. The somatic mutation was conditioned by a pair of recessive genes (prpr) controlling leaf color. Plants that possess this mutation manifest in their young and adult leaves, stipulae and stems with a purplish tint[71,72]. The high genetic proximity between 'Mundo Novo' and 'Purpuracea' is expected, because the two cultivars share almost identical parentage in the Bourbon-Typica group. 'Mundo Novo' was selected from the cross of 'Bourbon' x 'Sumatra', whereas 'Purpuracea' was selected from the cross of 'Maragogype' x 'Caturra'.

      Based on ∆K value computed by STRUCTURE HARVESTER, the 71 coffee accessions were partitioned into two clusters as the most probable number of K[50]. At a high assignment coefficient value (Q > 0.80), the first group (the blue bars in Fig. 4b) included 16 members, comprising the 14 SGAB genotypes, 'Mundo Novo' and 'Purpuracea'. This result is consistent with that depicted by the UPGMA tree, where 'Mundo Novo' and 'Purpuracea' grouped with the 14 SGAB genotypes and formed a tight sub-cluster. The second Bayesian cluster included 11 members (the orange bar in Fig. 4b), comprising 'Catimor', 'Sarchimor', and other accessions with introgressed background of 'Timor Hybrid'. The third category included 44 accessions that have an assignment coefficient (Q-Value) < 0.80, thus were classified as admixed genotypes between the first and the second clusters.

      The intra-cultivar variation of the phenotypic traits as mentioned above is compatible with several previous reports in other Arabica coffee cultivars. Using SSR and AFLP markers, Tran[26] observed that even the elite cultivars of Arabica coffee, which have been exposed to intensive selection, still show a certain degree of genetic variation amongst individuals within each cultivar. Steiger et al.[25] analyzed genetic diversity among Arabica cultivars using AFLP markers and observed within cultivar variation in 'Typica', 'Mokka Hybrid', 'Catuaí/Caturra' and 'Catimor'. Gichimu[29] reported the existence of a significant yield variation among 'Ruiru 11' siblings. Specifically, Tassone et al. [10] explored the intra-cultivar variation in 'Mundo Novo' and selected progenies with high yield capacity, bean size, plant height, upper and lower canopy diameter, stem diameter, and plant vigor. Among all the selected traits, they obtained the highest selection gain in bean size (the percentage of beans retained in the sieve was 17). This type of variation was attributed to residual heterozygosity and out-crossing in Arabica coffee[25]. Indeed, outcrossing is not uncommon in Arabica coffee although the species is generally considered as an inbreeding species[73]. In some cultivars, the outcrossing rate can reach as high as 76%[28]. However, the possibility of outcross was excluded, judged by the UPGMA clustering analysis based on SNP genotyping (Fig. 4a). Despite the significant phenotypic variations, all the 14 progenies formed a tight sub-cluster, which closely grouped with 'Mundo Novo'. This result demonstrates that all the 14 SGAB genotypes have a common parent of 'Mundo Novo'. The molecular variation among the 14 genotypes, therefore, could be attributed to residual heterozygosity or mutations. Nonetheless, whole genome resequencing will need to be applied to these progenies to further understand the genetic basis of the intra-cultivar variation relating to phenotypic traits, including cup quality attributes. Specifically, a sample panel with diverse variation of quality-related traits, need to be included in future studies for genome-wide association analysis.

    • Intra-cultivar variation in the progenies of a traditional coffee cultivar 'Mundo Novo 379/19' was explored to select genotypes with superior cup quality. A total of 14 SGAB coffee genotypes, selected from the putative progenies of 'Mundo Novo 379/19' were evaluated for sensory attributes, chemical compositions, and agronomic traits. SNP markers were applied to verify their relationship with the putative parent 'Mundo Novo', together with 57 other as other reference cultivars from the coffee collection of EPAMIG, Brazil. The six SGAB coffee genotypes evaluated all showed a high sensory quality (sensory scores > 80), which enabled them to be classified as specialty coffee by the SCA standard and in the category of very good and excellent. The top three genotypes (SGAB 1, SGAB 7 and SGAB 8) were selected for future genetic improvement programs. This study demonstrates a good example of selecting dual high-yield, higher-cup quality performance cultivar from existing traditional varieties, using the approach combining field experiment, cup quality assessment, chemical analysis and SNP marker-based molecular characterization.

      • We thank Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Instituto Brasileiro de Ciência e Tecnologia do Café (INCT/Café), under FAPEMIG grant (CAG APQ 03605/17). To Professor Vinicius Quintão Carneiro for the support in the statistical analysis of sensory and biochemical data. We would also like to give special thanks Stephen Pinney of USDA-ARS for SNP genotyping of the coffee samples.

      • Antonio Chalfun-Junior is the Editorial Board members of Journal Beverage Plant Research. He was blinded from reviewing or making decisions on the manuscript. The article was subject to the journal's standard procedures, with peer-review handled independently of these Editorial Board members and their research groups.

      • # These authors contributed equally: Gabriela Ester Ferraz, Marlon Enrique López

      • Supplemental Table S1 Coffee bean granulometry of 14 coffe genotypes and two controls (Catuaí Vermelho IAC 99 and Mundo Novo)- Year 2015. *Values represented in percentages.
      • Supplemental Table S2 Summary of genotypes parameters two by two comparison for the variables height, branch length, number of nodes per branch and stem diameter. The β0’s represent the initial values (intercepts) and the β1’s the increment of the values of the response variables over time (linear regression coefficients).
      • Supplemental Table S3 Results of regression analysis of the estimates of  β0’s (intercepts) and β1’s  (linear regression coefficients – slope) for the variables plant height, branch length, number of nodes and stem diameter in 14 genotypes of C. arabica with 95% confidence interval.
      • Supplemental Table S4 Result fo SNP genotyping on 14 'Mundo Novo' progenies and 57 reference accessions of Arabica coffee (Coffea arabica),  using nano-fluidic array.
      • Copyright: © 2023 by the author(s). Published by Maximum Academic Press, Fayetteville, GA. This article is an open access article distributed under Creative Commons Attribution License (CC BY 4.0), visit https://creativecommons.org/licenses/by/4.0/.
    Figure (4)  Table (2) References (73)
  • About this article
    Cite this article
    Ferraz GE, Lópes ME, Gonçalves FMA, Malta MR, Lima RRD, et al. 2023. Progeny selections of coffee cultivar 'Mundo Novo' with potential for the specialty coffee market. Beverage Plant Research 3:1 doi: 10.48130/BPR-2023-0001
    Ferraz GE, Lópes ME, Gonçalves FMA, Malta MR, Lima RRD, et al. 2023. Progeny selections of coffee cultivar 'Mundo Novo' with potential for the specialty coffee market. Beverage Plant Research 3:1 doi: 10.48130/BPR-2023-0001

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return