Professional Documents
Culture Documents
Artigo Cientifico Analise de Correlação
Artigo Cientifico Analise de Correlação
net/publication/322645164
Maturation periods for Coffea arabica cultivars and their implications for yield
and quality in Brazil
CITATIONS READS
3 353
5 authors, including:
Some of the authors of this publication are also working on these related projects:
PREVISÃO DO CUSTO DA PRODUÇÃO DE SOJA USANDO TÉCNICAS DE MODELAGEM COMPUTACIONAL View project
All content following this page was uploaded by Lucas Eduardo De Oliveira Aparecido on 21 February 2018.
Abstract
BACKGROUND: Climatic conditions directly affect the maturation period of coffee plantations, affecting yield and beverage
quality. The quality of coffee beverages is highly correlated with the length of fruit maturation, which is strongly influenced
by meteorological elements. The objective was to estimate the probable times of graining and maturation of the main
coffee varieties in Brazil and to quantify the influences of climate on coffee maturation. We used degree days to estimate
flowering/graining periods (green fruit) and flowering/maturation periods (cherry fruit) for all cultivars. We evaluated the
influence of climate on the time of maturity using Pearson correlation and nonlinear regression analysis and successfully
mapped the influences of these elements.
RESULTS: Arabica coffee matured up to 2–3 months earlier in São Paulo, where air temperatures (T AIR ) were higher, than
in Minas Gerais, which would allow earlier harvesting and the training of seedlings at the beginning of the rainy season.
Catuaí-Amarelo-IAC-62 cultivar needed 205–226 days between the end of flowering and maturation at locations with high T AIR
and 375–396 days at locations with low T AIR .
CONCLUSION: Water surplus and deficit were generally the most important variables for coffee maturation. Coffee matured
faster in regions with high T AIR and evapotranspiration, moderate altitudes and deficits. Acaiá-Cerrado-MG-1474 and
Icatu-Precoce-Amarelo-3282 were cultivars with an early cycle.
© 2018 Society of Chemical Industry
Figure 1. Main regions for coffee production in the states of Minas Gerais and São Paulo, Brazil.
Table 2. Degree days total subperiods flowering–graining Table 3. Pearson correlation coefficients (r) between matu-
and flowering–maturation of the main coffee varieties in Brazil ration time of coffee and the geographical elements (ALT, alti-
tude), water elements (WS, water surplus; WD, water deficit)
Sum of degree days (∘ )
and energetic elements (T AIR , air temperature; PET, evapo-
Σ Flowering– Σ Flowering– traspiration; PET/R, evapotraspiration/rainfall)
Number Cultivar graining maturation
Maturation ALT WS WD T AIR PET PET/R
C1 Acaiá Cerrado MG 1474 2009.4 2423
Maturation 1.000
C2 Icatu Precoce Amarelo 3282 2094.2 2440.9
ALT 0.677 1.000
C3 Mundo Novo 379-19 1987.8 2569.3
WS 0.903 0.683 1.000
C4 Catuaí Vermelho IAC 99 2133.2 2674.2
WD −0.902 −0.749 −0.721 1.000
C5 Topázio MG 1190 2373.8 2719
TAIR −0.717 −0.921 −0.773 0.786 1.000
C6 Catuaí Vermelho IAC 15 2496.2 2742.9
PET −0.718 −0.837 −0.785 0.829 0.965 1.000
C7 Catuaí Amarelo IAC 17 2423 2779.3
PET/R −0.838 −0.717 −0.926 0.873 0.845 0.914 1.000
C8 Rubí MG 1192 1987.8 2811.4
C9 Catuaí Vermelho IAC 144 2423 2811.4
C10 Catuaí Amarelo IAC 62 2423 2843.5
Qo
PET = 0.01 × × T × ND (1)
2.45
Qo = 37.6 × DR
[( ) ] Figure 2. Decreasing influence of the geographical elements (ALT, alti-
𝜋
× × hn × sin Φ × sin 𝛿 + cos Φ × cos 𝛿 × sin hn tude), water elements (WS, water surplus; WD, water deficit) and energetic
180 elements (T AIR , air temperature; PET, evapotraspiration; PET/R, evapotraspi-
(2) ration/rainfall) during coffee maturation.
( )
360 × JD
DR = 1 + 0.033 × cos (3) DEF = PET − AET (10)
365
[( ) ] {
𝛿 = 23.45 × sin
360
× (JD − 80) (4) 0, if WC < 0
SURi = (11)
365 (P − PET)i − ALTi , if WC = 0
hn = arcos [− tan Φ × tan 𝛿] (5) where SUR = water surplus at the soil–plant–atmosphere sys-
tem (mm); ALT = soil water storage of the current month −
where ND is the number of days in the month, JD is Julian day and soil water storage of the preceding month (mm); i = the monthly
Φ is altitude. period; PET = potential evapotranspiration (mm); P = rainfall (mm);
We evaluated the influence of hydric conditions on the matu- NAC = total rainfall − potential evapotranspiration; WC = available
ration period. Water deficit (WD) and surplus (WS) – components water capacity (mm); STO = soil water storage (mm); DEF = water
of the monthly climatic water balance – were calculated follow- deficiency at the soil–plant–atmosphere system (mm) and
ing the methodology of Thornthwaite and Mather,28 using a water AET = actual evapotranspiration (mm).
storage capacity of 100 mm: We used degree days (DD) (Eqn (12)) to estimate the flower-
ing/graining period (green fruit) and flowering/maturation period
{
NACi = NACi−1 + (P + PET)i (cherry fruit) for all cultivars in both states. DD was set for all cli-
if (P − PET)i < 0 = (NACi ) (6) matic series and was recorded daily from the start of flowering:
STOi = WCe WC
[ ]
{ ∑ Tmax − Tmin
STOi = (P − PET)i + STOi−1 DD = − Tbase (12)
if (P − PET)i ≥ 0 = (STOi ) (7) 2
NACi = WC ln WC
Figure 3. Water balance28 for São Paulo State: (A) Campos do Jordão; (B) São Carlos; (C) São Simão; and (D) Votuporanga. Water surplus (WS, mm mo−1 );
water deficit (WD, mm mo−1 ).
average near 18 September (with a standard deviation of 10 days) The regression analysis allowed us to verify the dependence of
in all regions. We therefore used 18 September as the average date the maturation period on the hydric, energetic and geographical
of flowering for all regions. elements. We used linear and nonlinear regression analysis with
The maturation time varies greatly from year to year. The use four parameters (logistics models) (Eqn (14)). Maturation time was
of the date of 18 September in this work is valid because it the dependent variable, and the other variables were the inde-
reflects the average of several scientific and unpublished works pendent variables. We used ordinary least squares for estimating
of coffee cooperatives in the region, which always indicate the 2∘ the nonlinear models, which minimizes the sum of squares of the
ten-day period of September as the most likely period of coffee model errors33 by optimizing the ‘generalized reduced gradient’34 :
maturation. Years with marked thermal and/or water stress will
ymax − ymin
probably influence these estimated dates; however, the objective y = ymax + ( )p (14)
of this work was an average evaluation of the maturation period 1 + xx
that is in agreement with previous studies.20–31 0
The DDs were retained from the flowering, and the dates of where y = the estimated cumulative variable; ymax = the maximum
graining and maturation were considered when the total DD point of the curve (amplitude); Y min = the minimum point of the
was reached. Thermal requirements quantified by Carvalho et al.32 curve; x 0 = the inflection point (X is value of maximum growth);
were used to determine the DDs for the main Brazilian coffee and p = maximum growth rate.
cultivars (Table 2). The geographic information system was used for spatial interpo-
We also evaluated the influence of the hydric (rainfall, water sur- lation of the probable dates of coffee graining and maturation for
plus and water deficit), energetic (air temperature and PET) and all locations using the method of kriging35 with a spherical model
geographical (altitude) elements on the time of maturity using with one neighbour and a spatial resolution of 0.25∘ (27.7 km).
Pearson correlation and nonlinear regression analysis. The uni-
variate Pearson correlation (Eqn (13)) analysis provided a more
detailed quantification of the relationships between coffee mat- RESULTS AND DISCUSSION
uration and the energetic, hydric and geographical elements. This
Coffee maturation correlated variously with the hydric, energetic
analysis was also important for understanding how each element
and geographical elements. Altitude (ALT) and water surplus (WS)
affected coffee maturation:
positively affected, and water deficit (WD), air temperature (T AIR ),
evapotranspiration (PET), and the PET/rainfall ratio (PET/R) nega-
n ( )
∑ tively affected the maturation time (Table 3). High T AIR , high PET
Yobsi − Yobs) × (Yelementsi − Yelements
i=1
rate, moderate ALT and WD thus contributed to faster coffee mat-
r= √
)2 √∑ (
uration. These results are consistent with previous studies,5–11
∑(
n
n
)2
which also reported that ALT and PET/R showed positive and neg-
Yobsi − Yobs × i=1
Yelementsi − Yelements
i=1 ative effects with coffee maturation, respectively.
(13) Pearson correlation coefficients were arranged in descending
where Yelements = energetic, water and geographical elements; order to identify variables with the most influence on the matu-
Yobsi = observed maturation; n = number of data; and k = number ration period. The hydric variables (WS and WD) were the most
of independent variables in the regression. important, with correlations of 0.903 and −0.902, respectively
Figure 4. Water balance28 for Minas Gerais State: (A) Araxá; (B) Barbacena; (C) Caldas; (D) Caparaó; (E) Juiz de Fora; (F) Lavras; (G) Machado; (H) São Lourenço;
(I) São Sebastião do Paraiso; and (J) Viçosa. Water surplus (WS, mm mo−1 ); water deficit (WD, mm mo−1 ).
(Fig. 2). ALT was less well correlated with maturation time, with a the soil–plant–atmosphere system.37 In the state of São Paulo,
coefficient of 0.677. These correlations indicated that the climatic WD was >25 mm mo−1 for April–October at São Simão and for
conditions of the environment had more influence than altitude April–November at Votuporanga, which promoted a reduction
on the time of maturity of the coffee crop. Moreto et al.36 also in the maturation time of coffee (Fig. 3). In Minas Gerais, WD was
observed that water conditions were the elements that most influ- highest in July and August at all locations. WD was highest at
enced the quality and maturation of orange fruits. Araxá, reaching 50 mm mo−1 in August, and lowest at Caldas and
We analysed the hydric elements for all locations, because São Lourenço in August, at <10 mm mo−1 (Fig. 4). In São Paulo
WS and WD correlated most strongly with maturation time. State, Campos do Jordão location had the lowest T AIR , around
Water balance is the most efficient indicator for characterizing 15 ∘ C in the period of May to August (Fig. 3). In Minas Gerais,
Figure 5. Annual water deficit (A) and water surplus (B) of locations producing coffee in the State of São Paulo and Minas Gerais, Brazil.
Figure 6. Estimation of flowering/maturation period (days) of coffee in function of (A) altitude (m), (B) water surplus (mm y−1 ), (C) water deficit (mm y−1 ),
(D) air temperature (∘ C y−1 ), (E) evapotranspiration (mm y−1 ) and (F) evapotranspiration/rainfall rate (mm y−1 ).
Figure 7. Likely month for the occurrence of granado and the maturation of coffee in São Paulo. CPJ, Campos do Jordão; SSM, São Simão; SCL, São Carlos;
VTP, Votuporanga. (A) Acaiá Cerrado MG-1474; (B) Icatu Precoce Amarelo-3282; (C) Mundo Novo 379-19; (D) Catuaí Vermelho IAC-99; (E) Topázio MG-1190;
(F) Catuaí Vermelho IAC-15; (G) Catuaí Amarelo IAC-17; (H) Rubí MG-1192; (I) Catuaí Vermelho IAC-144; and (J) Catuaí Amarelo IAC-62.
T AIR was highest in November and December and lowest for the crop cycle increased with ALT.13–39 Coffee maturation time in
May–August. Average T AIR was lowest (around 15 ∘ C) at Caldas our study decreased as WD, T AIR , PET and PET/R increased. A WD of
and highest (around 24 ∘ C) at Araxá (Fig. 4). 1200 mm y−1 (Fig. 6C), T AIR of 25 ∘ C (Fig. 6D) and PET of 1450 mm
WD has most influence on the development of agricultural y−1 (Fig. 6E) promoted a time from flowering to maturation of
crops.38 The cumulative WDs and WSs varied among the locations 200 days.
(Fig. 5). The annual average WD was 90.35 mm y−1 in São Paulo and The probable days of graining and maturation varied among
63.40 mm y−1 in Minas Gerais. Annual WD was highest in São Paulo locations and cultivars. Campos do Jordão–SP showed an average
at Votuporanga and in Minas Gerais at Araxá, reaching 231.40 and monthly T AIR < 15 ∘ C, and graining occurred in February for culti-
126.10 mm y−1 , respectively. WD was <20 mm y−1 in São Paulo at vars Acaiá Cerrado MG-1474, Icatu Precoce Amarelo 3282, Mundo
Campos do Jordão and in Minas Gerais at Caldas and São Lourenço. Novo 379-19, Catuaí Vermelho IAC-99 and Rubí MG-1192 (Fig. 7A,
Modelling the effects of meteorological elements on coffee mat- B, C, D, H) and in April for cultivars Topázio MG-1190, Catuaí Ver-
uration time resulted in different relationships. The accumulation melho IAC-15, Catuaí Amarelo IAC-17, Catuaí Vermelho IAC-144
of WS and increasing ALT had a sigmoidal relationship with the and Catuaí Amarelo IAC-62 (Fig. 7E, F, G, I, J).
time of coffee maturation. An ALT of 1550 m and WS of 950 mm Maturation occurred mostly in March to June for all cultivars, but
y−1 promoted the stabilization of maturation at 390 days (Fig. 6A, in places with high T AIR there is a reduction in the cycle of coffee,
B). ALT was an indicator of fruit ripening, because the length of as observed in VTP, SSM and SCL (Fig. 7).
Figure 8. Likely month for the occurrence of granado and the maturation of coffee in Minas Gerais. ARX, Araxá; BBC, Barbacena; CLD, Caldas; CPR, Caparaó;
LVR, Lavras; MCH, Machado; SLR, São Lourenço; SSP, São Sebastião Paraiso; JUF, Juiz de Fora; VIÇ, Viçosa. (A) Acaiá Cerrado MG-1474; (B) Icatu Precoce
Amarelo-3282; (C) Mundo Novo 379-19; (D) Catuaí Vermelho IAC-99; (E) Topázio MG-1190; (F) Catuaí Vermelho IAC-15; (G) Catuaí Amarelo IAC-17; (H) Rubí
MG-1192; (I) Catuaí Vermelho IAC-144; and (J) Catuaí Amarelo IAC-62.
Figure 9. Duration of graining season of the main coffee cultivars A) Acaiá Cerrado MG-1474; B) Icatu Precoce Amarelo-3282; C) Mundo Novo 379-19; D)
Catuaí Vermelho IAC-99; E) Topázio MG-1190; F) Catuaí Vermelho IAC-15; G) Catuaí Amarelo IAC-17; H) Rubí MG-1192; I) Catuaí Vermelho IAC-144 and, J)
Catuaí Amarelo IAC-62.
The periods of graining and maturation varied in Minas Gerais the amplitude of the phenological phases of the cultivars (Fig. 8).
(Fig. 8). Graining occurred on average between January and In these colder places, the quality of the drink is higher, because
April and maturation between May and July. The time between the fruits presented high amounts of caffeine, trigonelline and
flowering and maturation was shortest at Araxá, Lavras and chlorogenic acid.5–11
Machado, which detracts from fruit quality. Hydric conditions and temperatures varied widely in São Paulo
Average monthly T AIR is <15 ∘ C at Barbacena and Caldas,22 so the and Minas Gerais (Figs 3 and 4), so the coffee period from flowering
cultivars in our study formed grains between April and August and to maturation varied by up to 6 months among locations (Figs 7
matured between July and October, increasing the crop cycle and and 8). The time between flowering and fruit formation was
Figure 10. Duration of maturation season of the main coffee cultivars: (A) Acaiá Cerrado MG-1474; (B) Icatu Precoce Amarelo-3282; (C) Mundo Novo 379-19;
(D) Catuaí Vermelho IAC-99; (E) Topázio MG-1190; (F) Catuaí Vermelho IAC-15; (G) Catuaí Amarelo IAC-17; (H) Rubí MG-1192; (I) Catuaí Vermelho IAC-144;
and (J) Catuaí Amarelo IAC-62.
longer in the regions with milder T AIR , such as in the southern The flowering/graining period for the long-cycle cultivar Catuaí
and east-central regions of Minas Gerais. T AIR was highest in Vermelho IAC-15 varied between 179 and 197 days in the northern
São Paulo, especially in the north, with the longest maturation and northeastern regions of São Paulo and between 333 and
period (Fig. 3D) and annual WD accumulation (Fig. 5) on fewer days 352 days in the south of Minas Gerais. This result shows that the
between flowering and grain formation (Fig. 9). same cultivar presents distinct maturation in different places.
Cycle times were similar for the cultivars Catuaí Amarelo IAC-17, Cultivars with premature cycles, such as Mundo Novo 379-19
Catuaí IAC-144, Catuaí Amarelo IAC-62 and Catuaí IAC-15 (Fig. 9). and Catuaí Amarelo IAC-17, had flowering/graining durations of
147–160 days in the north of São Paulo and of 251–263 days in the altitude terroirs of Costa Rica, Orosi and Santa María de Dota. J Sci
southern region of Minas Gerais (Fig. 9). These two cultivars are the Food Agric 85:1869–1876 (2012).
6 Silva SA, Lima JSS and Alves AI, Spatial study of grain yield and
most commonly planted by the coffee producers in the region.39
percentage of bark of two varieties of Coffea arabica L. to the
Rapid grain formation is not optimal, because nutrient assimilation production of quality coffee. Biosci J 26:558–565 (2010).
and accumulation are not successful, which cause problems in 7 Martinez HEP, Poltronieri Y, Farah A and Perrone D, Zinc supplementa-
grain filling, less dense fruit, and even a lower quality of coffee tion, production and quality of coffee beans. Rev Ceres 60:293–299
beverage.17 (2013).
Faster maturing cultivars such as Acaiá Cerrado MG-1474 had 8 Silva AS, Queiroz DM, Pinto FAC and Santos NT, Characterization and
delimitation of the terroir coffee in plantations in the municipal
shorter periods between flowering and maturation in São Paulo, district of Araponga, Minas Gerais, Brazil. Rev Cienc Agron 45:18–26
mainly in the north, with a length between 174 and 193 days (2014).
(Fig. 10). Very early ripening produces poor coffee beverages, 9 Castro RD and Marraccini P, Cytology, biochemistry and molecu-
because chlorogenic acids and tryptophan have not fully devel- lar changes during coffee fruit development. Braz J Plant Physiol
oped, promoting bitter-tasting beverages.15 The harvest of very 18:175–199 (2006).
10 Gay C, Estrada F, Conde C, Eakin H and Villers L, Potential impacts of
early crops can also coincide with the end of the rainy season, climate change on agriculture: a case of study of coffee production
which is not ideal, because the high relative humidity favours in Veracruz, Mexico. Clim Change 79:259–288 (2006).
microorganisms that ferment coffee grain pulp, which also reduces 11 Barbosa JN, Borem FM, Cirillo MA, Malta MR, Alvarenga AA and Alves
the quality of the coffee beverage.40 HMR, Coffee quality and its interactions with environmental factors
The cultivars Rubi MG-1192 and Catuaí IAC-144 had flowering in Minas Gerais, Brazil. J Agric Sci 4:181–190 (2012).
12 Anzanello R, Souza PVD and Coelho PF, Thermal requirement, phenol-
periods similar to the maturation period of the late cultivar Catuaí ogy and productivity of Niagara Branca, Niagara Rosada and Con-
Amarelo IAC-62. Catuaí Amarelo IAC-62 needed 205–226 days cord cultivars submitted to two harvests by vegetative cycle. Rev
between the end of flowering and maturation at locations with Bras Frutic 34:366–376 (2012).
high T AIR (northern and northeastern São Paulo) and 375–396 days 13 Bardin-Camparotto L, Camargo MBP and Moraes JFL, Estimated time of
at locations with low T AIR (southern and east-central Minas Gerais) grain ripening of different Coffea arabica cultivars for the São Paulo
State, Brazil. Cienc Rural 42:594–599 (2012).
(Fig. 10). 14 Silva EA, Mazzafera P, Brunini O, Sakai E, Arruda FB, Mattoso LHC et al.,
A slow ripening of the coffee grains produces better coffee The influence of water management and environmental conditions
beverages, because more sugar can accumulate.13 The longest on the chemical composition and beverage quality of coffee beans.
period between flowering and maturation was in Minas Gerais, Braz J Plant Physiol 17:229–238 (2005).
mainly in the south, with durations between 324 and 343 days. 15 Fagan EB, Souza CHE, Pereira NMB and Machado VJ, Effect of time on
coffe bean (Coffea sp) grouth in cup quality [sic]. Biosci J 27:729–738
Moderate T AIR and WD and high WS are prevalent in the southern
(2011).
region of Minas Gerais (Fig. 4). 16 Angelico CL, Pimenta CJ, Chalfoun SM, Chagas SJR, Pereira MC and
Chalfoun Y, Different stages of maturity and times of bagging in
relation to quality of coffee. Coffee Sci 6:8–19 (2011).
CONCLUSIONS 17 Silva PA, Rabelo VM, Calixto JMR, Coelho PO and Gorski IRC, Quality
assessment of coffee grown in Campos Gerais, Minas Gerais State,
Coffee arabica L. matured up to 2–3 months earlier in São Paulo, Brazil. Acta Sci Technol 36:739–744 (2014).
where T AIR is higher, than in Minas Gerais. 18 López-Urrea R, Montoro A, Mañas F, López-Fuster P and Fereres E,
The climatic conditions in southern Minas Gerais allow a slow Evapotranspiration and crop coefficients from lysimeter measure-
ripening of the fruit. Water surplus and water deficit are the most ments of mature ‘Tempranillo’ wine grapes. Agric Water Manage
important variables for coffee maturation. 112:13–20 (2012).
19 Miranda MN and Campelo JH Jr, Accumulated heat unit for the
The fruit matures earlier where air temperature and evapotran-
bean cv. ‘Carioca’ sowing-maturation sub-period in Colorado
spiration are high, altitude is moderate and when even a small do Oeste, Rondônia State, Brazil. Pesq Agropec Trop 40:180–185
water deficit occurred. Low air temperature and water deficit have (2010).
larger effects on short-cycle cultivars. 20 Pezzopane JRM, Pedro Júnior MJ, Camargo MBP and Fazuoli LC, Heat
Acaiá Cerrado MG-1474 and Icatu Precoce Amarelo 3282 were requirements of Mundo Novo coffee for the flowering-harvest phe-
nological stage. Cienc Agrotecnol 32:1781–1786 (2008).
cultivars with an early cycle. Acaiá Cerrado MG-1474 had shorter
21 Ajayi AE and Olufayo AA, Evaluation of two temperature stress indices
periods between flowering and maturation in the north of São to estimate grain sorghum yield and evapotranspiration. Agron J
Paulo, with a length between 174 and 193 days. Catuaí Vermelho 96:1282–1287 (2004).
IAC-144 and Catuaí Amarelo IAC-62 were late cycle cultivars. 22 de Oliveira Aparecido LE, Rolim GS and Souza OS, Flowering and har-
vesting periods of macadamia–walnut for areas of the southeastern
coffee region. Rev Bras Frutic 36:170–178 (2014).
REFERENCES 23 Franklin J, Davis FW, Ikegami M, Syphard AD, Flint LE, Flint AL et al.,
1 Vaast P, Bertrand B, Perriot JJ, Guyot B and Gén M, Fruit thinning Modeling plant species distributions under future climates: how
and shade improve bean characteristics and beverage quality of fine scale do climate projections need to be? Global Change Biol
coffee (Coffea arabica L.) under optimal conditions. J Sci Food Agric 19:473–483 (2012).
86:197–204 (2006). 24 Roberts MJ, Schlenker W and Eyer J, Agronomic weather measures
2 Bertrand B, Boulanger R, Dussert S, Ribeyre F, Berthiot L, Descroix F in econometric models of crop yield with implications for climate
et al., Climatic factors directly impact the volatile organic compound change. Am J Agric Econ 95:236–243 (2012).
fingerprint in green Arabica coffee bean as well as coffee beverage 25 Camargo AP, Water balance in the state of São Paulo, 3rd edn. IAC,
quality. Food Chem 135:2575–2583 (2012). Campinas, p. 24 (1971).
3 Silva FM, Arré TJ, Tourino ES, Gomes TS and Alves MC, Use of ethrel 26 Thornthwaite CW, An approach towards a rational classification of
on the mechanized and selective coffee (Coffea arabica L.) harvest. climate. Geogr Rev 38:55–94 (1948).
Coffee Sci 4:178–182 (2009). 27 Camargo AP and Camargo MBP, An analytic revision of the potential
4 Handayani R, Coffee and its flavor. Rekayasa Pertanian Dan Biosistem evapotranspiration. Bragantia 59:125–137 (2000).
4:222–227 (2016). 28 Thornthwaite CW and Mather JR, The water balance. Laboratory of
5 Avelino J, Barboza B, Araya JC, Fonseca C, Davrieux F and Guyot B, Climatology, Drexel Institute of Technology: Centerton, AR. Publica-
Effects of slope exposure, altitude and yield on coffee quality in two tions in Climatology 8: (1955).
29 Lima EP and Silva EL, Base temperature, crop coefficients and 36 Moreto VB, Rolim GS, Zacarin BG and Latado RR, Agrometeorolog-
degrees-day for Arabic coffee plants in the implantation phase. R ical models for forecasting the qualitative attributes of ‘Valência’
Bras Eng Agric Ambiental 12:266–273 (2008). oranges. Theor Appl Climatol 130: 847–864 (2017).
30 Zacharias AO, Camargo MBP and Fazuoli LC, Agrometeorological 37 Davi H, Peer review report 2 On ‘Coupling a water balance model
model for estimating the beginning of the flowering period for cof- with forest inventory data to predict drought stress: the role of
fee crop (Coffea arabica L.). Bragantia 67:249–256 (2008). forest structural changes vs. climate changes’. Agric For Meteorol
31 Nunes FL, Camargo MBP, Fazuoli LC, Rolim GS and Pezzopane 217:14–16 (2016).
JRM, Agrometeorological models to estimate the duration of 38 Carvalho HP, Dourado Neto D, Teodoro REF and Melo B, Climatological
flowering–maturation stage for three Coffea arabica cultivars. hydric balance, effective soil water storage and transpiration in
Bragantia 69:1011–1018 (2010). coffee culture. Biosci J 27:221–229 (2011).
32 Carvalho HP, Camargo R, Gomes MWN and Souza MF, Classification of 39 Silva Sde A, de Queiroz DM, Ferreira WP, Corrêa PC and Rufino JL,
the development cycle of coffee cultivars by means of thermal sum. Mapping the potential beverage quality of coffee produced in the
Coffee Sci 9:237–244 (2014). Zona da Mata, Minas Gerais, Brazil. J Sci Food Agric 96:3098–3108
33 Draper NR and Smith H, Applied Regression Analysis (2nd edn). Wiley, (2015).
New York (1980). 40 Abrahão AA, Pereira RGFA, Borém FM, Rezende JC and Barbosa JC,
34 Lasdon LS and Waren AD, GRG2 user’s guide. Department of General Physical analyses and quimical composition of coffee submitted to
Business, School of Business Administration, University of Texas, different fungicidal treatments. Coffee Sci 4:100–109 (2009).
Austin, TX (1982).
35 Krige D, A statistical approach to some basic mine valuation problems
on the Witwatersrand. J Chem Metall Min Soc S Afr 52:119–139
(1951).