Relations between sensory quality and spectral indices in brazilian arabica coffees

Abstract

This article describes an investigation using spectral indices to characterize coffee production of Brazil, regarding beverage quality and possible correlations with the growing environment. The study evaluated 50 arabica coffee samples, 16 of which were natural process, and 34 were pulped coffes. These samples were originated from growing areas located in different altitude ranges and regions of the municipality, with similar planting spacing and predominance of Catuai cultivars. The samples were subjected to sensory analysis, which revealed that 58% of the samples were classified as specialty coffees: 3 natural, and 26 pulped coffes. Multiple correspondence analysis showed that average spectral indices, normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), and photochemical reflectance index (PRI), derived from images of the multispectral instrument (MSI), were not associated with the quality parameters of the coffee beverage. In contrast, the plant senescence reflectance index (PSRI) proved to be the relevant factor in the quality of the drink. In summary, the analysis of the relationship between the indices demonstrated that the NDVI, which measures the vegetative vigor of plants, showed an inverse correlation with the PSRI. Additionally, the principal component analysis suggested that samples collected from drier areas differed significantly from other geographic regions.

Key words
Coffea arabica; remote sensing; spatial variability; water stress

INTRODUCTION

The term “specialty coffee” connotes a group in which quality and the consumer’s experience are privileged. Specialty coffee quality is assessed using specific physical grading criteria and sensory, such as aroma and taste evaluation (Schuit et al. 2021, Tieghi et al. 2024). Elliot (2006) highlights that coffee connoisseurs use geography to illustrate both their knowledge and taste preferences. When ordering Sumatra, Kona, New Guinea Peaberry, Brazil Ipanema Bourbon, and similar specialty coffees, the connoisseur “orders a location for each cup” alluding to the importance of the geographical location and coffee quality.

The differentiation of coffees based on the notion of terroir allows the determination of potential areas to produce specialty coffees and the characterization of the coffee identity in these areas, exploiting the potential (Silva 2014). The definition of terroir must begin by identifying the differences in product quality. Then, other characteristics, such as microclimate, soil, and geography, are thoroughly evaluated, and the quality of the coffee beverage is evaluated (Naranjo et al. 2011, Silva 2014, Hervé 2020).

Improvements in satellite image technology and data processing have substantially increased the ability to map, quantify, and qualify land cover changes on a global scale (Rosa et al. 2021). Of the relevant information with application in agriculture are the spectral vegetation indices (VIs), defined as mathematical formulations developed from spectral data obtained by remote sensors, mainly from the red and near-infrared bands, which allow evaluations and estimates of the vegetation cover of an area. These mathematical transformations can be interpreted as semi-analytical measures of vegetation activity to detect spatial variability. Spectral vegetation indices (VIs) have been effectively used in vegetation monitoring and to analyze vegetation in responses to climate change, such as surface temperature variation (Abdul Athick et al. 2019, Liu et al. 2019).

The cloud-based platform Google Earth Engine (GEE) is an important tool for geospatial analysis on a planetary scale. This platform allows users to utilize Google’s computational resources to address high-impact issues, such as climate monitoring, disease tracking, deforestation, natural disasters, drought assessment, water management, and environmental protection (Gorelick et al. 2017).

The are research gaps on the relationship between water stress and coffee drink quality (Ahmed et al. 2021). Furthermore, no studies correlate spectral index data with the sensorial quality of coffee drinks. The fraction of absorbed photosynthetically active radiation, a crucial remote sensing model parameter of vegetation productivity (Liu et al. 2019).

Given the relevance of coffee cultivation and the current and future trends related to climate change, this work focuses on the increasing global demand and production of specialty coffees. In this context, the study of terroir, which shapes the unique characteristics of the beverage through the interaction between the plant’s genotype and the environment, becomes particularly relevant.

MATERIALS AND METHODS

Characterization of the study area

The study covers the limits of the municipality of Barra do Choça, located in southwestern Bahia, Brazil, within the coordinates 14°4’ to 15°46’ South and 40°24’ to 40°44’ West (Figure 1).

Figure 1
Location map of the municipality of Barra do Choça-BA. Datum: SIRGAS 2000 (IBGE 2005). Data base: IBGE 2006. Projection: UTM.

The municipality of Barra do Choça stands out as the highest coffee producer in the belt of Planalto Baiano. It ranks fourth in coffee production and first in planted areas in Bahia (IBGE 2017). Coffea arabica L. predominates in terms of land use in the municipality and socioeconomic relevance, emphasizing the Catuai varieties.

Figure 2 shows the study region’s climatic characteristics, characterized by a hot tropical climate (Köppen-Geiger), with mean annual temperatures of 18 °C, a rainy summer, and a dry winter lasting four to five months. The land has flat to gently undulating relief, with altitudes ranging from 384 to 1,056 m above sea level. The region’s biome is the Atlantic Forest, which hosts four types of vegetation: dense rainforest, seasonal semideciduous forest, seasonal deciduous forest, and a small region with pioneer formation. As one approaches the transition region with the Caatinga biome, it is observed that there are new deciduous vegetation patches. The predominant type of soil in the region is the dystrophic yellow Oxisol (IBGE 2006).

Figure 2
Thematic maps of the municipality of Barra do Choça-BA: vegetation, soils, climate, distribution of humidity/year and biome.

Coffee samples

Fifty batches of coffee beans (Coffea arabica L.) were used, 16 natural coffees (NC) and 34 pulped coffees (PC), from farms in Barra do Choça, Bahia, Brazil. The samples were deposited in the first coffee quality contest in the municipality, held by the Cooperativa Mista dos Cafeicultores de Barra do Choça and region (COOPERBAC) on September 2, 2022.

The contest included natural and pulped coffes categories, distinguished by the post-harvest process. Therefore, part of the grains was processed naturally (without removing the pericarp before drying); another part washed process or wet process (a technique widespread in the region), pulped natural process (mechanical removal of the pulp and epicarp before drying), and some batches were anaerobic fermentation process. All samples were submitted for drying.

Sensory analysis

The sensory panel was composed of five Q-grader tasters, following the Specialty Coffee Association (SCA) sensory analysis protocol, which assigns grades from 0 to 10 to the following attributes: fragrance/aroma, uniformity, presence of defects, sweetness, flavor, acidity, body, aftertaste, balance, and overall experience, with the latter aspect constituting the general impression (personal assessment of the taster), which totals a general score from 0 (zero) to 100 (one hundred) points.

In each evaluation, tests were carried out on three coded samples (blind test) representing each lot. The evaluation of the attributes of the samples proceeded according to the “cup test” standard (aspirate/taste/discard), and its concept of global quality was registered on the SCA’s standard Evaluation Form.

The coffees were classified based on their overall score as to the quality standard, referencing the classifications used by ABIC and SCA. For this study, samples with grades above 45 points were selected and classified according to Table I.

Table I
Adapted reference form for evaluating the global quality of the coffee beverage (ABIC 2023; SCA 2023). Lower quality coffees were discarded in this study.

SCA sensory analysis notes	Classification SCA	Classification ABIC	Classification used
< 45.00	Below standard	Inferior	-
45.01 – 59.99	Below standard	Traditional	1
60.00 – 71.99	Below standard	Superior	2
72.00 – 79.99	Below standard	Gourmet	3
80.00 – 100	Specialty	Gourmet	4

To better analyze the sensory results, the final scores were subdivided into classes to verify frequency and spatial quality. The first phase of the sensory analysis revealed 50 coffee samples with scores greater than 45 points, subdivided into four classes: Class 1 – traditional (45.00 – 59.99 points), Class 2 – superior (60.00 – 71.99 points), Class 3 – gourmet (72.00 – 79.99 points), and Class 4 – specialty (80.00 – 100 points).

Thematic mapping and generation of vegetation indices

Thematic maps of the total area of the municipality of Barra do Choça were generated. The vegetation indices were obtained across the entire municipality area; however, the analyzed values were extracted from points of the properties under study. The resources used in the development of this study are shown in Table II.

Table II
Resources used for image processing.

Resources	Description
Satellite images Source	MSI - Sentinel 2 (multispectral instrument) European Space Agency (ESA)
Software for image geoprocessing	Google Earth Engine Quantum Gis
Shapefiles	Municipal Network 2019 from SEI Bahia Vegetation IBGE Climate IBGE TOPODATA – Geomorphometric Database of Brazil – DSR/ INPE
Products	NDVI - Normalized Difference Vegetation Index (biomass index) EVI – Enhanced Vegetation Index (biomass index) PRI – Photochemical Reflectance Index (sunlight efficiency) PSRI – Plant Senescence Reflectance Index (stress indicator)
Time series	Sep 1, 2020, to Aug 30, 2022

The shapefile of the municipality of Barra do Choça was extracted from the cartographic database available on the IBGE website, from which thematic maps of climate, vegetation, biome, and distribution of humidity/year were generated. The topographic thematic map was prepared using the TOPODATA Project (INPE) image in GeoTIFF format, which is available on the INPE website.

The Google Earth Engine platform obtained the spectral indices scenes from the MSI sensor (Multispectral Instrument) mounted on the Sentinel-2 satellite. The sensor, with its thirteen spectral bands, four with a spatial resolution of 10 m, six with 20 m bands, and three with 60 m bands (Gómez 2017), was used to analyze spectral indices (Table III).

Table III
Vegetation indices equations.

Vegetation indices	Equation
NDVI	(b(‘B5’) - b(‘B4’)) / (b(‘B5’) + b(‘B4’))
EVI	(“2.5 * (b(‘B5’) - b(‘B4’)) / (b(‘B5’) + 6 * b(‘B4’) - 7.5 * b(‘B2’) + 1)
PRI	(b(‘B2’) - b(‘B3’)) / (b(‘B2’) + b(‘B3’))
PSRI	(b(‘B4’) - b(‘B3’)) / b(‘B6’))

The general method used digital image processing to extract spectral indices and obtain spectral characteristics associated with climate and vegetation cover for the time series from Sep 1, 2020, to Aug 30, 2022. The indices also correlated with the coffee quality produced in the Barra do Choça municipality. The chosen time series is analogous to the phenological period of coffee (Camargo & Camargo 2001).

Statistical analysis

The data were classified as follows: 1. post-harvest category: natural and pulped; 2. geographic data: latitude and longitude; 3. average spectral indices: NDVI, EVI, PRI, and PSRI; and 4. sensory quality scores. Principal Component Analysis (PCA) was conducted using the XLSTAT software.

A biplot graph was used to verify the correlation of spectral indices with sensory quality and spatial dependence.

RESULTS

Considering the samples analyzed in the first phase of sensory analysis, 29 obtained a score above 80 points, considered special, from which 26 were coffees in the pulped and 3 in the natural categories.

Figure 3 depicts that specialty coffees are not confined to a single region. However, they are instead dispersed across variable altitudes in three distinct regions: the South (1), the North (2), and the Northeast (3).

Figure 3
Hypsometric map of the municipality of Barra do Choça with location of analyzed coffees. Region 1- blue highlight, Region 2- orange highlight, Region 3- yellow highlight.

The regions were categorized based on altitude and vegetation criteria. The South Region (1) was identified by its semi-deciduous vegetation and higher altitudes, the North Region (2) by its medium altitudes and semi-deciduous vegetation, and the Northeast Region (3) by dense vegetation.

The South (1) and North (2) regions are particularly promising in producing good-quality natural coffees. These areas have less dense vegetation (Figure 3) due to lower rainfall, typical of seasonal deciduous and semi-deciduous forests, with cultivated areas at 900.01 – 1032.34 m altitudes. It was possible to observe the occurrence of special natural coffees (NC), particularly in the South region (1), which is situated in higher areas of the municipality.

The active variables, the final scores of sensory analyses, and means of the spectral indices (EVI, NDVI, PSRI, and PRI) obtained in the time series corresponding to phenological years I and II, applied in the analysis of principal components, are explained by 74.67% of the variability, 44.21% in PC1 and 30.46% in PC2 (Figure 4).

Figure 4
Biplot plot from principal component analysis (PCA). The points correspond to the observations: natural coffees (NC) and pulped coffees (PC). The vectors represent the active variables (scores and averages of the NDVI, PSRI, PRI and EVI spectral indices, in the period from Sep 1, 2020 to Aug 30, 2022).

The classic biplot analysis aims to group the samples (active observations) based on the similarities among the active variables. However, in Figure 4, the active observations, natural coffees (NC) and pulped coffees (PC), displayed a dispersed distribution across the Cartesian plane, preventing them from being differentiated by clusters in PC1 or PC2. It is assumed that no clusters were formed regarding the post-harvest treatment (NC and PC) in the multivariate analysis.

Variables located in the same quadrant have associations with each other. This observation makes it possible to infer that in negative PC1 values, the PSRI index and the final scores of sensory analyses are correlated. Similarly, the PRI, EVI, and NDVI indices were the variables that grouped the samples into positive PC1 values. It is also possible to affirm that the PSRI index has a negative correlation with the NDVI index, with a high degree of certainty, corroborating the studies previously carried out in Brazilian conditions by Volpato et al. (2013), who observed spatial relationships between NDVI and land surface temperature and precipitation data, where NDVI indices below 70% indicated the occurrence of water and/or temperature stress.

DISCUSSION

It can be observed that the spatial dispersion of the sensory quality among the coffees in the municipality did not result in distinct clusters (Figure 4). Special coffees were present at all altitude levels above 800 m, both in wetter regions and in drier areas, due to the post-harvest decision-making. However, a statistically significant association was established between quality and the PSRI spectral index, which indicates the presence of water stress, thus linking quality to drier areas.

Based on sensory and physical parameters, pulpe coffees stood out for their superiority, showing no interdependence spatial (with altitude or environmental conditions), as demonstrated by the dispersion of the samples (Figure 3). In contrast, natural coffees were predominantly found in higher altitude areas and lower humidity, indicating spatial dependence, as sensory quality tends to be consistent in higher altitude samples, especially in Region 1(south) and 2(north) (Figure 3). This result is corroborated by Martins et al. (2020), who state that pulped coffee processing is more suitable at lower altitudes, while at higher altitudes, both processes can be applied to produce beverages with distinct sensory profiles. It is also important to consider the cultural and technical dimensions involved, as pulped coffee producers have greater technical resources, while natural coffee producers in the region are entering the specialty coffee market and improving their post-harvest processes.

Several studies associate sensory quality of coffee with altitude and fruit maturation (Martins et al. 2020, Veloso et al. 2020, Ahmed et al. 2021, Guimarães et al 2019, Tolessa et al. 2017). In general, it is observed that regions with milder temperatures and higher altitudes have a longer cycle duration (the period between flowering and full maturation) (Bardin-Camparotto et al. 2012). The relationship between altitude and temperature is significant for tropical and subtropical regions, where a difference of a few hundred meters causes considerable environmental changes in biota adaptation and the consequent success of introducing species for agricultural cultivation (Fritzsons et al. 2016).

Altitude has a strong influence on changes in air temperature (Cargnelutti Filho et al. 2006). Lower temperatures delay fruit ripening, causing more chemical changes in the grains and a greater accumulation of sugars, certain acids, and amino acids, which enhance the drink’s characteristics (Vaast et al. 2005, 2006).

Contrary to what the literature often states, region 3, located in the Northeast portion of the municipality of Barra do Choça, at lower altitudes, 700-900m (Figure 5), is the most humid and rainy area with dense rainforest vegetation (Figure 3). Due to the environmental condition humid the harvest periods, the coffees from this region were, in their entirety, depulping by washed process and obtained high sensory scores.

The processing method contributes to variations in the coffee profile and its microbiota (Martinez et al. 2022a). The Fermentation in fruits can modify sensory attributes, in intensity and frequency, assigning special scores of aroma, fragrance, and acidity, as well as impairing quality, when not well (Góngora et al. 2024). Among the analyzed samples, ten underwent a controlled fermentation process; of these, 40% achieved scores above 80 points, 40% fell within the range of 72.00 to 79.99 points, and 20% obtained scores below 72 points, highlighting the effectiveness of the technique.

A justification for depulping coffees is based on reducing the area occupied by cement platforms (yards) and reducing drying time. This process optimizes mechanical dryers due to the removal of the pulp and reduces costs in the drying process (Borém & Reinato 2006). Continuously within the process, while the coffee is drying, fermentation occurs. Fermentation generates essential flavor compounds and enhances quality and sensory scores. However, many other mechanisms and interactions are still unknown, and the microbial population varies with the processing method (Martinez et al. 2022b).

The choice of the drying process is an important decision because delays in drying can promote unwanted fermentation, while excessive temperatures can also degrade the quality of the beverage. Studies conducted by Borém et al. (2008) and Taveira et al. (2015) identified that drying at temperatures above 60 °C negatively affects coffee quality. It was also observed that reducing sugars, total sugars, and sensory analysis scores decrease with increasing drying temperatures, regardless of the processing method used. Six farms employed a mixed drying process for the analyzed coffees, starting with an initial drying in greenhouses and finishing drying in a dryer. Four of these farms achieved scores above 80 points. Most of the properties utilized greenhouses for drying the coffee throughout the process due to their low cost and effective quality control.

Although the surrounding areas of Barra do Choça have a semi-arid climate, the seasonality characterized by rains concentrated in the summer and a drier winter makes the environment of the high areas of Bahia State peculiar and ideal for cultivating specialty coffees. However, it is known that the production of specialty coffees requires the coordinated and harmonious effort of the entire production chain, as the quality of the coffee is shaped from planting to post-harvest processing (Tolessa et al. 2017, Pereira et al. 2019). Additionally, terroir plays a fundamental role in coffee quality. Conceptually, terroir refers to a space where interactions between the physical and biological environment and agricultural practices occur, providing unique characteristics to the product originating from that space (Peng et al. 2020).

As all Brazilian coffee production is located in areas with latitudes greater than 4°, coffee is phenologically found in tropical conditions rather than equatorial ones. Therefore, it is possible to state that the cycle is well-defined: flowering occurs in spring, fruiting in summer, maturation in autumn, and harvest in winter (Camargo & Pereira 1994).

The schematization of the different phenological phases of the Arabica coffee tree is a reference for research and observations in coffee growing, allowing to identify the stages that require available soil moisture and the stages that a small water stress should take place to induce an abundant flowering; and enables, among other things, the understanding of the reproductive process and maturation predictability, according to the prevailing climatic variation (Camargo & Camargo 2001).

Coffea arabica cultivation in the state of Bahia, Brazil, has gained recognition for its quality, particularly with coffees from Planalto Baiano and high productivity in western Bahia, evidenced by success in national and international quality competitions. Despite this evidence, several agroclimatic zoning studies for coffee cultivation in Brazil underestimate the belts of Arabica coffee in Bahia, classifying them as agroclimatic zones of risk for coffee production (ZARC) due to the predominant conditions in their surroundings, characterized by a semi-arid. This situation arises from the need for more meteorological stations that represent coffee-growing areas, which limits discussions agrometeorological aspects of the environment.

The preparation of coffees through wet process from ripe fruits, with rapid elimination of the source of fermentation, can result in high-quality coffees if processed correctly (Borém 2023). There are general differences that affect the sensory attributes of “wet” and “dry” processed coffees, which cater to different market segments: the former tends to have a milder flavor, less body, and higher acidity, while the latter exhibits more body, astringency, and less acidity (Chalfoun & Fernandes 2013).

The NDVI spectral index is utilized in biomass estimates and changes in the phenological development of crops and can also be applied in studies involving hydrological variables, such as precipitation (Nezlin et al. 2005). During data collection, it was observed that the municipality of Barra do Choça has NDVI indexes indicating water stress compared to traditional coffee regions in Brazil.

Examining the correlation betweem the PSRI index and scores, using the PC1 biplot (Figure 4), reveals that the relationship remains unclear, as the score variable is also influenced by PC2, which accounts for 30.46% of the variation. Dessalegn & Landau (2008) report that caffeine biosynthesis and its accumulation in raw grains may be more pronounced during stress than under favorable conditions. In some works, there are citations of higher levels of caffeine in samples of high-quality Arabica coffee, when compared with other samples of Arabica with lower sensory quality (Farah et al. 2006, Franca et al. 2005). In coffee cultivation, regions located in areas with low humidity in the winter and concentrated rainfall in the summer feature productions of high sensory quality, such as the regions of Chapada Diamantina of Bahia and Cerrado of Minas Gerais.

In the present study, PSRI indicates that the coffee belt of Barra do Choça, a producer of quality coffee, was subjected to water stress. High yields and uniform flowering are induced by water stress in coffee growing regions. However, there are not many studies correlating sensory quality to the PSRI index or water stress. Matiello et al. (2006) states that, when the post-harvest period is very dry, several groups (series) of flower buds reach maturation and then go into dormancy more uniformly, leading to flowering and, consequently, more even fruiting and maturation, which facilitates the harvest and the quality of the coffee. Castro & Marraccini (2006) state that the highest values of sucrose were observed in regimes with greater water restriction. Sucrose is the main representative of free sugars in mature arabica coffee beans (Salva & Lima 2007). Sucrose accumulated in ripe fruits occurs in the final stages of maturation and constitutes one of the components for the beverage’s quality characteristics.

The EVI and PRI indices are correlated with score, as both are located in positive values of PC2, in the same quadrant, closer to the “Y” axis. The PRI index is increasingly used as an indicator of photosynthetic efficiency, related to the efficient use of sunlight. According to Valeriano (2008), the direction of sun exposure plays an important role in evapotranspiration and in water balance of crops. In addition, slope exposure influences air temperature, affecting the duration of the production cycle and determining the harvest time, which is important for the quality of coffee beans (Ferreira et al. 2012).

CONCLUSIONS

The NDVI index, which reflects plant vigor, shows no relationship with quality and exhibits a negative correlation with the average PSRI index values. On the other hand, the PSRI index, used as an indicator ofbo water stress, is correlated with quality but still requires further studies, together with EVI and PRI. Additionally, the results suggest that post-harvest treatment had a significant impact on the variation in the quality of coffee from the Barra do Choça region in Bahia, Brazil.

REFERENCES

• ABDUL ATHICK ASM, SHANKAR K & NAQVI HR. 2019. Data on time series analysis of land surface temperature variation in response to vegetation indices in twelve Wereda of Ethiopia using mono window, split window algorithm and spectral radiance model. Data Brief 9: 104773. Doi: 10.1016/j.dib.2019.104773. PMID: 31763418; PMCID: PMC6864355.
• ABIC - ASSOCIAÇÃO BRASILEIRA DAS INDÚSTRIAS DE CAFÉ. 2023. Programa de Qualidade do Café. Available at: https://www.abic.com.br/certificacoes/qualidade/ Accessed Feb 20, 2023.
  » https://www.abic.com.br/certificacoes/qualidade/
• AHMED S ET AL. 2021. Climate Change and Coffee Quality: Systematic Review on the Effects of Environmental and Management Variation on Secondary Metabolites and Sensory Attributes of Coffea arabica and Coffea canephora. Front Plant Sci 12: 708013. https://doi.org/10.3389/fpls.2021.708013.
  » https://doi.org/10.3389/fpls.2021.708013
• AYOADE JO. 2004. Introdução à climatologia para os trópicos. 10 Ed., Rio de Janeiro, Brasil.
• BARDIN-CAMPAROTTO L, CAMARGO MBP & MORAES JFL. 2012. Probable ripening time for different Arabica coffee cultivars for the State of São Paulo. Ciência Rural 4: 594-599. (In Portuguese, with abstract in English). https://doi.org/10.1590/S0103-84782012000400003.
  » https://doi.org/10.1590/S0103-84782012000400003
• BORÉM FM. 2023. Tecnologia pós-colheita e qualidade de cafés especiais. 1 Ed., Universidade Federal de Lavras, Lavras, MG, Brasil.
• BORÉM FM, MARQUES ER & ALVES E. 2008. Ultrastructural analysis of drying damagein in parchment Arabica coffee endosperm cells. Biosyst Eng 99: 62. doi:10.1016-66.
• BORÉM FM & REINATO CHR. 2006. Quality of pulped coffee subjected to different drying processes = Qualidade do café despolpado submetidos a diferentes processos de secagem. Revista Brasileira de Armazenamento 9: 26-31.
• CAMARGO AP. 1985. Florescimento e frutificação de café arábica nas diferentes regiões cafeeiras do Brasil. Pesquisa Agropecuária Brasileira 20: 831-839.
• CAMARGO AP & CAMARGO MBP. 2001. Definition and outline for the phenological phases of arabic coffee under brazilian tropical conditions = Definição e esquematização das fases fenológicas do cafeeiro arábica nas condições tropicais do Brasil. Bragantia, Campinas 60: 65-68. https://doi.org/10.1590/S0006-87052001000100008.
  » https://doi.org/10.1590/S0006-87052001000100008
• CAMARGO APDE & PEREIRA AR. 1994. Agrometeorology of the coffee crop. Geneva: World Meteorological Organization. World Meteorological Organization. CAgM Report, 58.
• CARGNELUTTI FILHO A, MALUF JRT, MATZENAUER R & STOLZ AP. 2006. Altitude e coordenadas geográficas na estimativa da temperatura mínima média decendial do ar no estado do Rio Grande do Sul. Pesquisa Agropecuária Brasileira 41: 893-901. https://doi.org/10.1590/S0100-204X2006000600001.
  » https://doi.org/10.1590/S0100-204X2006000600001
• CASTRO RD & MARRACCINI P. 2006. Biochemistry and molecular changes during coffee fruit development. Brazil J Plant Physiol 18: 175-199. https://doi.org/10.1590/S1677-04202006000100013.
  » https://doi.org/10.1590/S1677-04202006000100013
• CHALFOUN SM & FERNANDES AP. 2013. Efeitos da fermentação na qualidade da bebida do café. Visão Agrícola, 105-108.
• DESSALEGN B & LANDAU B. 2008. More than meets the eye: The role of language in binding and maintaining feature conjunctions. Psychol Sci 19: 189-195. doi: 10.1111/j.1467-9280.2008.02066.x.
• ELLIOTT C. 2006. Considering the connoisseur: Probing the language of taste. Canadian Review of American Studies 36: 229-236.
• FARAH A, MONTEIRO MC, CALADO V, FRANCA A & TRUGO LC. 2006. Correlation between cup quality and chemical attributes of Brazilian coffee. Food Chemistry 98: 373-380. https://doi.org/10.1016/j.foodchem.2005.07.032.
  » https://doi.org/10.1016/j.foodchem.2005.07.032
• FERREIRA, WPM, RIBEIRO MF, FERNANDES FILHO EI, SOUZA CF & CASTRO CCR. 2012. As características térmicas das faces noruega e soalheira como fatores determinantes do clima para a cafeicultura de montanha. Documentos - Embrapa Café, Brasília, 10: 34.
• FRANCA AS, MENDONÇA JCF & OLIVEIRA SD. 2005. Composition of green roasted coffees of different cup qualities. Food Science and Technology 38: 709-715. https://doi.org/10.1016/j.lwt.2004.08.014.
  » https://doi.org/10.1016/j.lwt.2004.08.014
• FRITZSONS E, MANTOVANI LE & WREGE MS. 2016. Relação entre altitude e temperatura: Uma contribuição ao zoneamento climático no Estado de Santa Catarina, Brasil. Revista Brasileira de Climatologia. https://doi.org/10.5380/abclima.v18i0.39471
  » https://doi.org/10.5380/abclima.v18i0.39471
• GÓNGORA CE, HOLGUÍN-STERLING L, PEDRAZA-CLAROS B, PÉREZ-SALINAS R, ORTIZ A & NAVARRO-ESCALANTE L. 2024. Metataxonomic Identification of Microorganisms during the Coffee Fermentation Process in Colombian Farms (Cesar Department). Foods (Basel, Switzerland) 13(6): 839. https://doi.org/10.3390/foods13060839.
  » https://doi.org/10.3390/foods13060839
• GÓMEZ MGC. 2017. Joint use of Sentinel-1 and Sentinel-2 for land cover classification: A machine learning approach. Lund University GEM thesis series.
• GORELICK N, HANCHER M, DIXON M, ILYUSHCHENKO S, THAU D & MOORE R. 2017. Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sensing of Environment 202: 18-27. https://doi.org/10.1016/j.rse.2017.06.031.
  » https://doi.org/10.1016/j.rse.2017.06.031
• GUIMARÃES RJ, BORÉM FM, SHULER J, FARAH A & ROMERO JCP. 2019. Coffee Growing and Post-harvest Processing. In: FARAH A (Ed), Coffee: Production, Quality and Chemistry, 1st ed., London - England: Royal Society of Chemistry 1: 26-88.
• HERVÉ A. 2020. Wine Yeast Terroir: Separating the Wheat from the Chaff-for an Open Debate. Microorganisms 8: 787. https://doi.org/10.3390/microorganisms8050787.
  » https://doi.org/10.3390/microorganisms8050787
• IBGE - INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICAS. 2005. Manual técnico de geodésia. 3a ed., Rio de Janeiro: IBGE. Available: https://www.ibge.gov.br
  » https://www.ibge.gov.br
• IBGE - INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICAS. 2006. Recursos Naturais. Base de dados espacial 2006. Informações por Cidade e Estado. Available: https://www.ibge.gov.br/geociencias/cartas-e-mapas/informacoes-ambientais/15829-solos.html Accessed: 10 Aug 2022.
  » https://www.ibge.gov.br/geociencias/cartas-e-mapas/informacoes-ambientais/15829-solos.html
• IBGE - INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICAS. 2017. Censo Agro. Rio de Janeiro. Available: https://censoagro2017.ibge.gov.br/templates/censo_agro/resultadosagro/agricultura.html
  » https://censoagro2017.ibge.gov.br/templates/censo_agro/resultadosagro/agricultura.html
• LIU EH, ZHOU GS & ZHOU L. 2019. Fraction of absorbed photosynthetically active radiation over summer maize canopy estimated by hyperspectral remote sensing under different drought conditions. Ying Yong Sheng Tai Xue Bao 30(6): 2021-2029. Chinese. doi: 10.13287/j.1001-9332.201906.041. PMID: 31257775.
• MARTINEZ SJ, BATISTA NN, BRESSANI APP, DIAS DR & SCHWAN RF. 2022b. Molecular, Chemical, and Sensory Attributes Fingerprinting of Self-Induced Anaerobic Fermented Coffees from Different Altitudes and Processing Methods. Foods (Basel, Switzerland) 11: 3945. https://doi.org/10.3390/foods11243945.
  » https://doi.org/10.3390/foods11243945
• MARTINEZ SJ, BRESSANI APP, SIMÃO JBP, PYLRO VS, DIAS DR & SCHWAN RF. 2022a. Dominant microbial communities and biochemical profile of pulped natural fermented coffees growing in different altitudes. Food Research International (Ottawa, Ont.) 159: 111605. https://doi.org/10.1016/j.foodres.2022.111605.
  » https://doi.org/10.1016/j.foodres.2022.111605
• MARTINS PMM, BATISTA NN, MIGUEL MGDCP, SIMÃO JBP, SOARES JR & SCHWAN RF. 2020. Coffee growing altitude influences the microbiota, chemical compounds and the quality of fermented coffees. Food Res Int 129: 108872. doi: 10.1016/j.foodres.2019.108872. Epub 2019 Dec 6. PMID: 32036899.
• MATIELLO AWR. 2006. Estresso ou não meu cafezal. Revista brasileira de tecnologia cafeeira: Coffea, Varginha-MG, III, 10: 29-30.
• NARANJO RDP ET AL. 2011. Fingerprints for Main Varieties of Argentinean Wines: Terroir Differentiation by Inorganic, Organic, and Stable Isotopic Analyses Coupled to Chemometrics. J Agric Food Chem 59: 7854-7865.
• NARANJO M, VÉLEZ L & ROJANO B. 2011. Antioxidant activity of different grades of Colombian coffee. Rev Cuba Plantas Med 16: 164-173.
• NEZLIN NP, KOSTIANOY AG & LI B. 2005. Inter-annual variability and interaction of remote-sensed vegetation index and atmospheric precipitation in the Aral Sea region. J Arid Environ 62: 677-700. https://doi.org/10.1016/j.jaridenv.2005.01.015.
  » https://doi.org/10.1016/j.jaridenv.2005.01.015
• PENG Y, ROELL EY, MØLLER AB, ADHIKARI K, BEUCHER A, GREVE MB & GREVE MH. 2020. Identifying and mapping terrons in Denmark. Geoderma. https://doi.org/10.1016/j.geoderma.2020.114174
  » https://doi.org/10.1016/j.geoderma.2020.114174
• ROSA MR ET AL. 2021. Hidden destruction of older forests threatens Brazil’s Atlantic Forest and hallenges restoration programs. Science Advances 7: 1. Doi:10.1126/sciadv.abc4547.
• SALVA TJG & LIMA VB. 2007. A composição química do café e as características da bebida e do grão. O Agrônomo. Campinas 59: 57-59.
• SCA - SPECIALTY COFFEE ASSOCIATION. 2023. Protocols: coffee cupping standards.
• SCHUIT P, MOAT J, GOLE TW, CHALLA ZK, TORZ J, MACATONIA S, CRUZ G & DAVIS AP. 2021. The potential for income improvement and biodiversity conservation via specialty coffee in Ethiopia. PeerJ 9: e10621. Doi: 10.7717/peerj.10621. PMID: 33614264; PMCID: PMC7879952.
• SILVA SA, LIMA JSS & ALVES AI. 2010. Spatial study of grain yield and percentage of bark of two varieties of coffea arabica l. to the production of quality coffee = Estudo espacial do rendimento de grãos e porcentagem de casca de duas variedades de Coffea arabica L. visando a produção de café de qualidade. Bioscience Journal 26: 558-565.
• SILVA SA, QUEIROZ DM, PINTO FAC & SANTOS NT. 2014. Characterization and delimitation of the terroir coffee in plantations in the municipal district of Araponga, Minas Gerais, Brazil. Rev Ciênc Agron 45: 18-26.
• TAVEIRA JHDS, BORÉM FM, ROSA SDVF, OLIVEIRA PD, GIOMO GS, ISQUIERDO EP & FORTUNATO VA. 2015. Post-harvest effects on beverage quality and physiological performance of coffee beans. Afr J Agric Res 10: 1457-1466. Doi:10.5897/AJAR2014.9263.
• TIEGHI H ET AL. 2024. Effects of geographical origin and post-harvesting processing on the bioactive compounds and sensory quality of Brazilian specialty coffee beans. Food Res Int 186: 114346. doi: 10.1016/j.foodres.2024.114346. Epub 2024 Apr 19. PMID: 38729720.
• TOLESSA K, D’HEER J, DUCHATEAU L & BOECKX P. 2017. Influence of growing altitude, shade and harvest period on quality and biochemical composition of Ethiopian specialty coffee. J Sci Food Agric 97(9): 2849-2857. doi: 10.1002/jsfa.8114. Epub 2016 Dec 9. PMID: 27786361.
• VAAST P, ANGRAND J, FRANCK N, DAUZAT J & GÉNARD M. 2005. Fruit load and branch ring-barking affect carbon allocation and photosynthesis of leaf and fruit of Coffea arabica in the field. Tree Physiology 25: 753-760. doi: 10.1093/treephys/25.6.753.
• VAAST P, 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. J Sci Food Agric 86: 197-204. http://dx.doi.org/10.1002/jsfa.2338.
  » https://doi.org/10.1002/jsfa.2338
• VALERIANO MM. 2008. Topodata: guia para utilização de dados geomorfométricos locais. São José dos Campos: INPE. Available at: http://www.dsr.inpe.br/topodata/data/guia-enx.pdf Accessed: Feb 13, 2023.
  » http://www.dsr.inpe.br/topodata/data/guia-enx.pdf
• VELOSO TGR, DA SILVA MCS, CARDOSO WS, GUARÇONI RC, KASUYA MCM & PEREIRA LL. 2020. Effects of environmental factors on microbiota of fruits and soil of Coffea arabica in Brazil. Sci Rep 10: 14692. doi: 10.1038/s41598-020-71309-y. PMID: 32895415; PMCID: PMC7477199.
• VOLPATO MML, VIEIRA TGC, ALVES HMR & SANTOS WJR. 2013. Modis images for for agrometeorological monitoring of coffee areas. Coffee Science 2: 176-182.

Publication Dates

• Publication in this collection
  17 Mar 2025
• Date of issue
  2025

History

• Received
  21 Aug 2024
• Accepted
  21 Nov 2024