Exposure To Climate Risk: A Case Study For Coffee Farming In The Region Of Alta Mogiana, São Paulo

Abstract

Studies around the world show an increase in global average temperatures, with a consequent increase in extreme events and changes in the distribution of precipitation, causing a decrease in agricultural production and changes in planting areas. This study analyzed the exposure to climate risk that the coffee crop in the region of Alta Mogiana/SP, Brazil has been presenting in the past thirty years (1991-2021). Time series of daily data of maximum and minimum temperature and precipitation were used. By the statistical tests we observed a trend of increase in maximum temperatures daily of approximately 1.4°C and minimum daily of 0.8°C in the municipalities of the region and a trend towards a decrease in precipitation of 0.9 mm daily, indicating greater exposure of the coffee crop in the region to climate risk and increased vulnerability for the coffee producer. In view of these analyses, a literature review was carried out, suggesting agroforestry systems and mechanical irrigation as the most promising strategies to manage climate risk in coffee plantations. In addition, drought-resistant cultivars, training courses for farmers, increased rural insurance, and nutritional control of the plants can also be considered efficient options for climate exposure in coffee plantations from Alta Mogiana.

Key words
Agriculture; climate change; climate risk management; coffee tree; Mann-Kendall

INTRODUCTION

Agriculture is the dominant form of land use globally, involving important economic, social, and cultural activities and providing a wide range of ecosystem services ( Shiferaw et al. 2014 ). However, due toits nature, agriculture remains highly sensitive to climaticvariations ( Shiferaw et al.2014 ) because plants require specific climaticconditions in the phenological stages of plant development, forexample, flowering or ripening of fruits. Thus, one can say that“despite technological advances over the past few decades,agricultural production continues to depend on weather and climate”( Santos et al.2018 ). About 80% of the variability of agriculturalproduction relates to atmospheric conditions during its cycle,since farmers cannot control natural phenomena ( Assunção & Wander 2014 , Heinemannet al. 2017 , Hoogenboom 2000 ).

In the ${21}^{st}$ century, discussions on the relationship between climate andagriculture have gained great relevance in the context of climateand environmental change. In 2018, the Special Report of theIntergovernmental Panel on Climate Change (IPCC) warned about theimpacts of global warming of 1.5C above pre-industrial levels.These changes can cause environmental and economic damage invarious sectors, compromising development and social security (IPCC2018, 2021) and increasing the vulnerability of certain socialgroups. Direct impacts of climate change on agricultural activitycan affect income and production, among other aspects, and causechanges in geographic distribution, with sectoral and regionalrepercussions in several economic sectors, thus compromising foodsecurity.

In this context, coffee is one of the most important cash cropsfor the economy of South American countries, with a production in2019 reaching 32 million tons (FAO 2019). Brazil stands out as thelargest producer and exporter of coffee and in 2019, the equivalentof 45 million dollars of this product was exported (FAO 2019). Inthe Southeast of the country, the states of Minas Gerais, EspíritoSanto, and São Paulo accounted for the largest productions in thecountry in the 2019/2020 biennium (EMBRAPA 2020).

For Brazil, studies ( Assad& Pinto 2008 , CEDEPLAR & FIOCRUZ 2008 , Marengo et al. 2009 , Schaeffer et al.2008 ) show the negative consequences of climatechange, since Brazilian agribusiness has production and exportvolumes that contribute to the nutrition of several countries, withthis sector being one of the most important for the Brazilian tradebalance. In a more recent scenario, it has been observed thatseveral natural systems are being affected by regional climatechange ( Resende et al.2019 ). It is assumed that impact from these changeswill be greater in the agricultural sector, especially in countriesthat have agriculture as their primary economic activity ( Porfirio et al. 2018 ),such as Brazil ( Resende etal. 2019 , Tavares et al. 2018 ).

The Arabica coffee crop in Brazil is economically viable underspecific climatic conditions, with average annual temperaturesbetween 18C and 23C and ideal annual temperatures between 19C and 21C, average annual rainfall levels between 1.200 mm and 1.800 mm,and altitudes between$400m$and$1.200m$( Thomaziello et al.2000 ). In addition, rainfall must also have adequateintensity/distribution to promote plant phenological developmentand crop management (Coffe & Climate 2015). Due to theimportance of temperature for the flowering of the coffee bean,only the months of September and October were analyzed regardingthe maximum temperature of the series. According to Gornall et al. 2010 ,extremes of temperature can be decisive in the growth of this crop,especially when they coincide with the main reproductive orvegetative phases of the coffee tree.

In the state of São Paulo, several studies show that climatechange is taking place. An increase in the frequency of hotter days( Dufek & Ambrizzi2008 ) and a greater number of days with heavy rains( Marengo & Camargo2009 ) are some of the examples of changes in weatherpatterns already identified in the state.

The Alta Mogiana region, located in the West of São Paulo state,has the largest volume of coffee exports (CCCMG 2017). Studiescarried out for Franca and Mococa, municipalities that make up theMogiana Average, showed evidence of change in the rainfall regime,which started to arrive in late October, when the expected month isSeptember ( Torres et al.2020 ). These changes, therefore, can jointly affectthe amount of water available in the soil ( Martins et al. 2019 ), the patterns ofevapotranspiration and water balance ( Assad et al. 2004 , Martins et al. 2018 , Tanasijevic et al.2014 ). All these changes will harm the sustainabilityof agricultural systems, as well as the areas suited to crops( Assad et al.2004 , Santos etal. 2017 ). For coffee, Rodrigues & Reis 2014 point out thatclimate is considered the main influencing factor in the productiveperformance of the plant and, thus, in the formation of productioncosts.

It is estimated that more than 80% of coffee-producing farms inthe country are family establishments with low adaptive capacity toadverse climatic events, making national coffee production morevulnerable to the effects of climate change, since Brazilian familyagriculture accounts for about 37% of that production ( Tavares et al. 2018 ).Coffee growing is an activity that highly depends on climaticconditions and is therefore vulnerable to the effects of climatechange. For this reason, studies that seek to discuss ways toincrease the adaptive capacity and resilience of agriculturalsystems are extremely important. When deepening works on therelationship between climate change and society in coffeeproduction, as is the case of this study, one must considerconcepts such as vulnerability, sensitivity, adaptive capacity, andexposure, so as to contribute to climate risk management.

Thus, vulnerability can be understood as the propensity of agiven population/locality to be affected by climate change due tothree fundamental elements: sensitivity, adaptive capacity andexposure (IPCC 2007). Sensitivity refers to how a system can beaffected, adversely or not, while adaptive capacity is related tothe ability to reduce or prevent damage from the exploitation ofbeneficial opportunities existing in systems ( Obermaier & Rosa2013 ). Exposure concerns the presence of people,systems, and their relationships, which can be adversely affectedby climate change (IPCC 2007, Obermaier & Rosa 2013).

Due to the complexity surrounding these concepts, indicators andindexes have been widely used in view of their ability tosystematize the collection of information and facilitate thevisualization of complex phenomena ( Menezes et al. 2018 , Quintão et al. 2017 ).Nevertheless, for the construction of these indicators, one mustunderstand the different factors associated with vulnerability,sensitivity, adaptive capacity, and exposure of the studiedlocations. Authors have been dedicating themselves to understandinghow the combination of biophysical, social, geographic, andeconomic factors may contribute to shaping the risks andsusceptibility of populations to climate change. Therefore, thereis a set of possible approaches (social, risk-hazard, socialecological) ( Adger2006 , Cutter& Finch 2008 , Füssel 2007 ).

The understanding of how much a given location is susceptible tochanges in climate dynamics enables the development of climate riskmanagement strategies as an alternative for agricultural producersto adapt, including alternatives for cultivation that will allowthem to maintain the quality and productivity of their crops. Thus,the study of climate variables as a means to assess climateconditions is an important tool in climate risk management( Kath et al.2021 ).

Based on the assumption that climate change can cause damage tocoffee production, this study aimed to analyze, by statisticaltests, the exposure to climate risk to which the Alta Mogianacoffee-producing region, in São Paulo, is susceptible. It proposes,based on a literature review, alternative strategies to be adopted by farmers as a way of managing climate risk in the region.

MATERIALS AND METHODS

Study area

The study region called “Alta Mogiana Paulista” is known for concentrating municipalities in the Northwest of São Paulo with altitude above$800m$, distinguishing itself in the state of São Paulo mainly by the production of excellent quality coffee due to favorable climatic conditions for growing this bean ( Faleiros et al. 2020 ).

The following municipalities stand out: Altinópolis, Batatais, Buritizal, Cajurú, Cristais Paulista, Franca, ltirapuã, Jeriquara, Nuporanga, Patrocínio Paulista, Pedregulho, Restinga, Ribeirão Corrente, Santo Antônio da Alegria, and São José da Bela Vista, as shown in Figure 1 .

Figure 1
Location of the study area – Alta Mogiana region in the state of São Paulo – Brazil.

According to the Köppen climate classification, the Alta Mogiana region has a tropical alternating rainy season climate (Aw), with rainy summers and dry winters ( Alvares et al. 2013 ). For Dubreuil et al. 2019 , if one had to choose a typical “Brazilian” climate, it would probably be this one.

Meteorological data

For this study, daily meteorological data were used, provided by the NASA Power Project, available on www.power.larc.NASA.gov. This NASA project provides a set of surface meteorological and solar radiation data estimated from information and models ( Sayago et al. 2020 ). The NASA’s data are widely used in agricultural modeling to understand, for example, the development of agricultural crops ( Bai et al. 2010 , Van Wart et al. 2015 ), cultivation simulations ( Ojeda et al. 2017 ), and modeling of planting-related diseases ( Savary et al. 2012 ).

NASA Power data were initially produced on a global grid of$$\frac{1}{2}$$degree by$$\frac{2}{3}$$degree and then crosslinked via bilinear interpolation to a global grid of half degree of longitude by half degree of arc latitude ( Sayago et al. 2020 ). It should be noted that the meteorological data provided by the NASA Power Project are freely available for download ( Sparks 2018 ), expanding this possibility for scientific studies.

In the study area, data were obtained from 6 different grids, which were automatically generated by NASA Power, taking into account the spatial distribution of each municipality in the State of São Paulo. In this case, smallest municipalities were framed together in the same grid, while the largest occupied a single grid, totalizing 15 municipalities that make up the region delimited as Alta Mogiana Paulista.

The 15 municipalities were divided as follows: Group 1: Altinópolis, Cajuru, and Santo Antônio da Alegria; Group 2: Batatais, Nuporanga, and São José da Bela Vista; Group 3: Buritizal, Jeriquara, and Ribeirão Corrente; Group 4: Cristais Paulista; Group 5: Franca, Itirapuã, Patrocínio Paulista, and Restinga; Group 6: Pedregulho. Daily data of maximum temperature, minimum temperature, and precipitation were used, considering the period from January 1991 to January 2021, generating a historical series of 30 years, which is the recommended period for climate analysis according to the World Meteorological Organization ( WMO 2019 ).

Statistical analyses were performed in two stages, which in the first stage daily maximum/minimum temperature and precipitation data were considered for the evaluation of the full period (1991 to 2021). In the second, daily maximum temperature data was considered in the months of September and October (for the full time series), since these months are considered decisive for the growth of the plant as they are marked by the flowering of the coffee bean. For both stages, the data were analyzed for understanding whether there are positive or negative trends that could affect coffee trees. Statistical tests were executed taking into account the climatic and environmental conditions recommended for the proper development of the coffee tree.

Validation of the meteorological data

The data from the NASA Power Project model were validated, comparing the maximum and minimum temperature, as well as precipitation, with data from Xavier et al. 2016 , which also refer to a set of meteorological data in high resolution ($0.25\times 0.25$) grids and which are freely available. The data from Xavier et al. 2016 were considered real data because they are field observations, obtained by meteorological stations and homogenized using advanced interpolation techniques.

The statistical procedure is the most important tool to investigate the quality of fit and accuracy of estimated data compared to measured data ( Aboelkhair et al. 2019 ). Thus, two statistical indexes were applied and executed between the data from NASA and from Xavier et al. 2016 .

These statistical measures are Root Mean Square Error (RMSE) (1) and Mean Bias (MB) (2).

$$RMSE=\sum _{i=1}^n\frac{{(O_i-P_i)}^2}{n}$$ (1)

$$MB=\sum _{i=1}^n\frac{(P_i-O_i)}{n}$$ (2)

It should be noted that in the equations presented,$P_i$is the predicted value (NASA Power model) and$O_i$is the observed value (values by Xavier et al. 2016 ).

RMSE represents the average standard deviation of the model’s prediction compared to the observation ( Bai et al. 2010 , Kobayashi & Salam 2000 ), and it is important to highlight, in this case, that the differences between prediction and observation are squared, so that this index shows the greatest deviations between one and the other. MB has an equal dimension between predicted and observed values, and its value is interpreted as the mean deviation between the samples ( Yaghoubi et al. 2020 ).

Many studies employ mean square error (MSE) and its rooted variant (RMSE), or mean absolute error (MAE) and its percentage variant (MAPE). Although they are useful, these rates share a common drawback: since their values can range between zero and$+\infty$, a single value of them does not give much information about the performance of the regression with respect to the distribution of the ground truth elements ( Yaghoubi et al. 2020 ).

Statistical tests

The tests for trend analysis in monthly time series aim to build models and analyses by non-parametric tests, being widely used in meteorology to assess the periodicity of phenomena ( Morettin & Toloi 2006 ).

According to Neves 2012 , in non-parametric tests, the original values are replaced by a rank order of values to calculate their statistics and are independent of the probability distribution of the studied data. Thus, non-parametric data are recommended for detecting trends in climatological data by the original characteristic of these data, which do not have a normal frequency and present positive asymmetries ( Sonali & Nagesh Kumar 2013 ).

The non-parametric test proposed by Mann 1945 and later adapted by Kendall 1975 verifies the value of the historical series with the other values, always following a sequential ordering process, counting the number of times the remaining terms are greater than the analyzed value.

Therefore, it is based on the rejection or acceptance of a null hypothesis ($H_0$), giving it the ability to deny or confirm the existence of a trend in the analyzed historical series with a certain significance level (95%).

The Mann-Kendall test, as it is known, can only be applied if the series is serially independent. Thus, the observations of the series are tested to understand whether they are independent and identically distributed, that is, the hypotheses are tested considering that: Hypothesis 0 ($H_0$): The observations in the series are independent and evenly distributed (there is no trend); Hypothesis 1 ($H_1$): The observations in the series have a trend over time (there is a trend).

The variable$S$of the Mann-Kendall test can be obtained by Equations (3) and (4) ( Hirsch & Slack 1984 ):

$$S=\sum _{c=1}^{n-1}\sum _{d=c+1}^{n}sign(x_d-x_c)$$ (3)

Considering that:

$$sign(x_d-x_c)=\{\begin{array}{l}+1,x_d>x_c\\ 0,x_d=x_c\\ -1,x_d<x_c\end{array}$$ (4)

Where$x_c$,$x_d$represent the data points at position$c$and$d$, respectively, and$n$is the size of the data series. The significance level,$Z$, can be calculated by Equation (5):

$$Z=\{\begin{array}{l}S-\frac{1}{\sqrt{var(S)}},S>0\\ 0,S<0\\ \frac{S+1}{\sqrt{var(S)}},S=0\end{array}$$ (5)

In this case,$t_c$accumulates for$t$and$c$denotes the iteration times. In this study, the confidence level was set at 95%.

The Mann-Kendall test is widely used to detect significant trends in meteorological data series, since this test compares the relative importance of sample data, which gives it the advantage of not requiring normalized distribution. Another advantage is its low sensitivity to abrupt interruptions in series ( Tabari et al. 2011 ). If there is a trend observed by the Mann-Kendall test, the Pettitt 1979 test ( 1979 ) is applied to characterize abrupt changes in a time series, without restrictions on the probability distribution ( Zhang & Lu 2009 ). The statistical method is calculated by Equation (6):

$$U_{t,r}=\sum _{i=1}^t\sum _{j=t+1}^{T}sign(x_i-x_j),1\le t<T$$ (6)

In this sense,$x_i$and$x_j$are data points in the time series of$T$.$U_{t,T}$, in turn, represents the statistical variable. Thus, the possible abrupt change point$K_t$can be calculated by Equation (7):

$$K_t=max\left|U_{t,T}\right|$$ (7)

Meanwhile, the corresponding significance probability$p$associated with$K_t$is defined by Equation (8):

$$p=2exp(\frac{-6K_t^2}{T^3+T^2})$$ (8)

Literature review

To analyze alternative strategies published in the literature that could be adopted by farmers as a means of climate risk management in the region, a literature review. In this sense, the following keywords was used: “Climate Risk Management Coffee”, “Adaptation and Mitigation Coffee” and “Climate Change Coffee” in the Scopus and SciELO databases, where articles published in national and international journals from 2011 to 2021 were selected, totalizing 13 studies in 13 different journals. The studies report research experiences that have shown success in climate change mitigation and adaptation for coffee producing areas. From this research, a summary table was created with the main studies and their respective intervention proposals.

RESULTS AND DISCUSSION

Validation of meteorological data

The results of the RMSE and MB indexes, presented in Figures 2 , 3 and 4 , showed that the NASA Power model was able to represent the maximum and minimum temperature, as well as the precipitation of the study area.

Figure 2
RMSE (a) and MB (b) values for the study area - maximum temperature (°C).

Figure 3
RMSE (a) and MB (b) values for the study area - minimum temperature (°C).

Figure 4
RMSE (a) and MB (b) values for the study area - precipitation (mm).

The RMSE values for maximum temperature were, on average, 2.125ºC (Figure 2a), with the highest value recorded at 2.448ºC (group 1) and the lowest at 1.738ºC (group 5). The MB values for maximum temperature were, on average, 0.840C (Figure 2b), lower than those observed in the model by Xavier et al. 2016 . The highest recorded value was 0.994ºC (group 3) and the lowest was −1.224ºC (group 1).

For minimum temperature, the average RMSE was 1.396ºC(Figure 3a), with the highest value recorded at 1.532ºC(group 2) and the lowest at 1.242ºC(group 3). In turn, MB presented a value of 0.183ºC(Figure 3b), higher than the data from Xavier et al. 2016 , on average. The highest recorded value was 0.808ºC(group 6) and the lowest was − 1.118ºC(group 4).

For precipitation, RMSE presented average values of 7.714 mm (Figure 4a), with the highest value at 13.487 mm (group 6) and the lowest at 4.888 mm (group 4). MB had an average value of -0.129 mm (Figure 4b); the highest value was 2.210 mm (group 6) and the lowest was -2.826 mm (group 5).

Despite the variability found between the predicted and observed values, the model was able to reproduce the average behavior of the variables under analysis, as well as their seasonality, indicating the possibility of using data from the NASA Power model for climate studies in this area.

In this sense, it is understood that the model reproduced seasonality satisfactorily, as the values of the analyzed parameters, mainly maximum and minimum temperature, were close to the observed values, showing that in the period of 30 years, the model was able to highlight the different temperatures at different times of the year.

Statistical tests

Table I presents the results from the Mann-Kendall test and Pettitt test regarding the maximum temperature historical series, and Table II shows the results from the minimum temperature historical series. Table III refers to the statistical analysis of precipitation data. All analyzed data had a p-value of < 0.0001 , and Sen’s slope of 95% indicating high data reliability, including the analysis of trends for the months of coffee flowering presented in Figure 5.

The results of the Mann-Kendall test for daily maximum temperature data in the study area pointed to an increasing trend throughout the analyzed time series (Rejects$H_0$). For the 6 groups studied, there was, according to the Pettitt test, a break in the data series with an abrupt increase in maximum temperatures in 2012 for the municipality of Pedregulho and in 2013 for the other municipalities that make up the region of study. The average temperatures before and after the series break year can also be seen in Table I. The average temperatures for the municipalities in groups 2, 3 and 4 showed an increase of 1.2°C from 2012 (group 4) and 2013 (groups 2 and 3), representing an increase of more than 1.0°C in 23 years of analysis for groups 2 and 3 and 22 years for group 4. In turn, the municipalities of groups 1 and 4 showed an increase in average daily temperature of 1.3°C for the same period of years.

Table I
Mann-Kendall test and Pettitt test values for maximum temperature.

Municipalities	Group	Kendall’s Tau	Interpretation	Temperature before the break year (°C)	Break year	Temperature after the break year (°C)
Altinópolis, Cajuru, and Santo Antônio da Alegria	1	0.111	Rejects H0 (there is a positive trend)	27.7	2013	29.0
Batatais, Nuporanga, and São José da Bela Vista	2	0.100	Rejects H0 (there is a positive trend)	28.5	2013	29.7
Buritizal, Jeriquara, and Ribeirão Corrente	3	0.107	Rejects H0 (there is a positive trend)	28.1	2013	29.3
Cristais Paulista	4	0.111	Rejects H0 (there is a positive trend)	26.9	2012	28.1
Franca, Itirapuã, Patrocínio Paulista, and Restinga	5	0.109	Rejects H0 (there is a positive trend)	27.4	2013	28.7
Pedregulho	6	0.106	Rejects H0 (there is a positive trend)	24.8	2012	26.2

Table II
Mann-Kendall test and Pettitt test values for minimum temperature.

Municipalities	Group	Kendall’s Tau	Interpretation	Temperature before the break year (°C)	Break year	Temperature after the break year (°C)
Altinópolis, Cajuru, and Santo Antônio da Alegria	1	0.061	Rejects H0 (there is a positive trend)	16.3	2012	17.0
Batatais, Nuporanga, and São José da Bela Vista	2	0.061	Rejects H0 (there is a positive trend)	17.2	2012	17.9
Buritizal, Jeriquara, and Ribeirão Corrente	3	0.062	Rejects H0 (there is a positive trend)	17.2	2012	18.0
Cristais Paulista	4	0.062	Rejects H0 (thereis a positive trend)	16.3	2012	17.0
Franca, Itirapuã, Patrocínio Paulista, and Restinga	5	0.063	Rejects H0 (there is a positive trend)	16.5	2012	17.3
Pedregulho	6	0.039	Rejects H0 (there is a positive trend)	14.1	2012	14.7

Table III
Mann-Kendall test and Pettitt test values for precipitation.

Municipalities	Group	Kendall’s Tau	Interpretation	Precipitation before the break year (mm)	Break year	Precipitation after the break year (mm)
Altinópolis, Cajuru, and Santo Antônio da Alegria	1	$-$0.048	Rejects H0 (there is a negative trend)	4.1	2005	3.3
Batatais, Nuporanga, and São José da Bela Vista	2	$-$0.043	Rejects H0 (there is a negative trend)	4.2	2007	3.3
Buritizal, Jeriquara, and Ribeirão Corrente	3	$-$0.045	Rejects H0 (there is a negative trend)	4.3	2007	3.5
Cristais Paulista	4	$-$0.058	Rejects H0 (there is a negative trend)	4.3	2007	3.5
Franca, Itirapuã, Patrocínio Paulista, and Restinga	5	$-$0.053	Rejects H0 (there is a negative trend)	4.2	2007	3.4
Pedregulho	6	$-$0.080	Rejects H0 (there is a negative trend)	4.4	2004	3.6

It is possible to highlight again the municipality of Pedregulho as the one that presented, within the study area, the greatest difference in daily average temperature after the break year (about 1.4°C). However, it should be noted that Pedregulho is located in group 6, the same one that presented the greatest difference between the data from Xavier et al. 2016 and from NASA. Thus, there are uncertainties regarding the magnitude of this result, although the p-value of < 0.0001 and Sen’s slope of 95% are significant values.

According to the Mann-Kendall test, the minimum temperature series of Alta Mogiana (Table II ) also showed an increasing trend. Based on the Pettitt test, a break in the data series, with an increase in minimum temperatures as of 2012 was observed in all municipalities in the region. Unlike the results found in Table I for maximum temperatures before and after the break year, minimum temperatures before the break year (2012) did not present differences greater than 1.0°C daily compared to temperatures after the break year. The greatest differences were recorded in groups 3 and 5 (0.8°C daily average), which correspond to the municipalities of Buritizal, Jeriquara, Ribeirão Corrente, Franca, Itirapuã, Patrocínio Paulista, and Restinga. Groups 1, 2 and 4 showed an increase of 0.7°C in 22 years while group 6 showed the smallest increase (0.6°C) within the same period.

Unlike the maximum/minimum temperature variables analyzed, the Mann-Kendall test for the precipitation series (Table III ) presented a negative trend, thus indicating a decrease in daily rainfall in this study area. The Pettitt test showed a break in the data series with a decrease in rainfall in 2004 for group 6, in 2005 for group 1, and in 2007 for the other groups. The greatest difference in rainfall for the data before and after the break year was observed in the municipalities from group 2 (Batatais, Nuporanga, and São José da Bela Vista), where a daily average difference of 0.9 mm was observed. The other municipalities showed differences of 0.8 mm, which represents a difference higher than 0.5 mm of daily precipitation in 17 years of analysis for groups 3, 4 and 5.

From Tables I , II , and III , it is important to emphasize Kendall’s Tau, which assumed values between$-1$and$+1$, with a positive correlation indicating that the classifications of both variables increase together (for maximum and minimum temperature). It is observed also that exist a negative correlation observed in the precipitation series indicating that as the classification of one variable increases, the other decreases ( Karmeshu 2012 ). The tests carried out, therefore, suggest that in the past 30 years there has been an increase in daily average minimum and maximum temperatures in the Alta Mogiana region, as well as a drop in daily average rainfall in the region.

Analysis of trends for coffee flowering months

Figure 5 presents the results of the analyses carried out for maximum temperatures of coffee flowering months. All the municipalities presented a positive trend (Rejects $H_0$). Analyzing only the maximum temperatures in the months of September and October in Alta Mogiana, an even more significant increase trend was noticed compared to the total set of maximum temperature data presented in Table I. The Pettitt test indicated a break in the temperature series, with an increase as of 2011 for group 6 and from 2013 for the other groups. Temperatures before and after the break year for group 6 showed a difference of 2.4°C, the highest compared to the other groups. The other groups showed an increase of 1.9°C (groups 1, 2 and 5), 2.0°C (group 4), and 2.1°C (group 3), representing an average increase of 1.9 in 23 years for flowering months.

Figure 5
Mann-Kendall test and Pettitt test values for maximum temperature (months of September and October).

It should be noted in this case that the analysis of flowering months suggests that Pedregulho is the most sensitive area to drought and has been more exposed to climate variability since 2011, prior to the period described in Table I , which dated 2012 as the series break year. According to Jaramillo et al. 2011 , increase in temperature can cause changes in the location of coffee plantations, as well as an increase in pests of this plant in several areas worldwide, such as Hypothenemus Hampei, better known as “Coffee Drill.” For Alfonsi et al. 2019 these changes also lead to an increase in typical coffee diseases, such as rust.

The Alta Mogiana region still has great economic viability for coffee production ( Goes & Chinelato 2018 ), but its profitability may be harmed by changes in weather patterns observed in the tables presented. For Petek et al. 2009 , the combination of temperature and precipitation affects the determination of suitable areas for cultivation. On the one hand, water stress reduces the plant’s thermal need, and, on the other hand, excess water requires an increase in temperature to complete the phenological stages, which are decisive for the productivity of the coffee tree ( Petek et al. 2009 ).

A fundamental condition is the dry period of about three months for the induction of the floral bud of coffee plants, which cannot be long, followed by a period of rain to start anthesis, which cannot be long either so as not to affect the fruiting stage. In addition, is necessary mild temperatures, since high temperatures cause physiological changes such as flower absorption (Coffe & Climate 2015). However, the increased drought period observed for the Alta Mogiana region is a harmful factor to Arabica coffee plants, because with a longer dry season period there is damage to the phenological phases of flower bud induction and of the floral bud ( B et al. 2010 ), impairing the final quality of the product.

These observations reinforce the fact that the Arabica coffee crop needs specific and adequate climatic conditions for its development. In a situation of reduced water availability, the coffee plant presents symptoms of leaf wilt, leaf fall, senescence of branches, induced deficiency of nutrients, and root death ( Assad et al. 2000 ). In some regions, it may be necessary to use irrigation ( Mesquita et al. 2016 ). The water deficit condition accelerates the maturation of Arabica coffee fruits, which impairs the development of the different physiological stages of seed formation and later their germination capacity ( Petek et al. 2009 ).

Due to climate change, coffee crop are more susceptible to phenological damage, which will consequently affect farmers. Lower productivity combined with a drop in product quality places them in an economically vulnerable situation. Thus, techniques aimed at managing climate risk are necessary as a means to mitigate financial losses and reduce the vulnerability of Alta Mogiana coffee producers.

Climate risk management for the Alta Mogiana region

In the case of coffee growing, climate risk management can be done at all stages of production, from the selection of beans for planting to post-harvest procedures. For this study, climate risk management can be understood as activities, initiatives, and proposals that contribute to the adaptation of coffee plantations to changes in the region’s climate dynamics (Coffe & Climate 2015).

First of all, the use of cultivars that are more resistant to drought events is one of the solutions with the greatest potential for adapting coffee production to climate change ( Tavares et al. 2018 ). In addition, cultivars that are more resistant to pests and diseases will be necessary, since the increase in temperature and the decrease in the amount of precipitation also contribute to the increase in pests and diseases inherent to the coffee tree. Tavares et al. 2018 point out that the interactive or complementary use of agricultural practices with the use of cultivars that are more tolerant to heat and drought can contribute to climate adaptation and, thus, to the sustainability of coffee growing even in areas where there will be climate restrictions in the future.

Another way to increase farmers’ resilience is by providing them more information about climate change and ways in which they can adapt (Coffe & Climate 2015). Improving access to early warning systems, such as local climate maps and expert committees, adopting of adaptation as part of the local development strategy, strength farmers’ organizations to facilitate and improve access to climate information and other support services (training, investment credit, agricultural insurance, etc.), are ways to prepare producers for future adverse scenarios (Coffe & Climate 2015). One of the examples of strategies for managing climate risk to coffee can be found in Southeastern Minas Gerais, where measures have been adopted focusing on production diversification, off-farm income diversification, increased access to irrigation, and expanded climate-related agricultural insurance ( Koh et al. 2020 ).

The offer of courses to expand climate education for farmers also has positive results. In Australia, for example, the offer of courses related to climate and agriculture showed results such as increased knowledge of farmers regarding meteorological monitoring and climatic effects on production, and greater awareness of future problems related to climate change ( George et al. 2009 ).

Nevertheless, studies carried out in different locations in Brazil ( Gonçalves et al. 2021 , Hernandes et al. 2004 ) and in countries such as Mexico ( Lin 2006 ), Nicaragua ( López-Sampson et al. 2020 ), and Colombia ( De Leijster et al. 2021 ) point to the feasibility of using trees intercropped with coffee plantations as an option to reduce climate impacts on the coffee tree. With adequate shading intensity, intercropping can produce larger fruits and increase the productive capacity of coffee trees ( Hernandes et al. 2004 ). Cultivation intercropped with tree crops can increase the amount of phytomass on the soil surface, offering protection against the impact of raindrops and avoiding sudden variations in humidity and temperature, in addition to being directly linked to the development of microbial communities, which are capable of indicating the level of soil degradation ( Alvarenga & Martins 2004 ). In other words, as pointed out by De Leijster et al. 2021 , the afforestation of coffee plantations can offer a wide range of long-term ecosystem services. In this context, we can mention an increase in insects such as bees, responsible for crop pollination ( Souza & Halak 2012 ) and a high potential for carbon sequestration, as pointed out by Gonçalves et al. 2021 . Thus, for the Alta Mogiana region, agroforestry systems have a great potential to contribute to climate risk management by providing, among other benefits, thermal protection. Since the statistical tests pointed to a trend of temperature increase over the past 30 years in the region, this is of great importance.

In addition to these initiatives, other proposals have been studied as an option for managing climate risk in coffee farming, significantly contributing to expanding the mitigation and adaptation possibilities that farmers will have to adopt in the near future. The wide range of possibilities for managing climate risk is also needed due to the considerable variety of geographical conditions of crops around the world ( Pinheiro et al. 2021 ). Table IV compiles some of the most promising initiatives for adaptation to climate change tested in different coffee producing areas in Brazil and worldwide.

Table IV
Scientific articles showing proposals for climate risk management for coffee growing.

Authors and year	Country where the study was carried out	Main topic addressed	Main conclusions
Hernandes et al. 2004	Brazil	Analysis of solar radiation indices in an agroforestry system	Agroforestry systems can provide thermal comfort to coffee plantations, in addition to increasing production and fruit size.
Sasaki et al. 2013	Brazil	Efficiency of electrostatic spraying on coffee plantations.	Savings in the use of agrochemicals by electrostatic spraying (37% efficiency) can provide a financial margin for investing in climate risk management initiatives.
Gonçalves et al. 2021	Brazil	Carbon sequestration in agroforestry systems	Agroforestry systems have a high potential for carbon sequestration, also contributing to climate change mitigation.
Lopes et al. 2021	Brazil	Smart irrigation	Feasibility of the project on Smart Irrigation of coffee plantations using smartphones for monitoring, contributing to water control in the plantation.
Tavares et al. 2018	Brazil	Consequences of climate change for coffee growing in Brazil	Use of heat and drought-resistant cultivars combined with innovative agricultural practices can minimize the impacts of climate change.
De Leijster et al. 2021	Colombia	Ecosystem services	Agroforestry systems are great enablers of long-term ecosystem services, contributing to environmental preservation.
Koh et al. 2020	Brazil	Climate risk for coffee production in Brazil	Diversification of production/income of the property and expansion of rural climate insurance, protecting the coffee producer from extreme weather events.
Ramírez-Builes & Küsters 2021	Colombia	Plant nutritional control	Nutrition with calcium and potassium showed an increase in plant nutrition. The plant presented greater resilience to biotic (pests) and abiotic (environment/climate) factors.
Tran et al. 2021	Vietnam	Efficiency of irrigation systems	Irrigation systems are more efficient when accompanied by public policies for rural credit and labor qualification. The technique is promising for locations with changes in the standard precipitation distribution.
Varona & Zayas 2016	Cuba	Analysis of the best cost-benefit in coffee irrigation	Irrigation of 52% of the area every 5 days presented the best cost-benefit. The efficient use of water contributes to lower production costs, making room for investments in climate risk management initiatives.
López-Sampson et al. 2020	Nicaragua	Effects of shading on coffee plantation	Agroforestry systems have shown significant benefits to coffee plantations, such as protection from high temperatures that can be a consequence of heat waves.
George et al. 2009	Australia	Results of applying a course focused on climate education for farmers.	The application of courses aimed at climate education for farmers increases their understanding of the subject.
Lin 2006	Mexico	Analysis of indicators of agroforestry systems against climatic events	Agroforestry systems have great potential to reduce the impacts of extreme temperatures on coffee plantations.

The initiatives presented in Table IV have great potential for success in the Alta Mogiana region. The increase in temperatures and the drop in rainfall observed in statistical tests will require farmers to employ new techniques and technologies for coffee crops. Thus, the use of technology, such as irrigation, will be essential to meet the water needs of crops. For Lopes et al. 2021 , irrigation in crops can be done in a practical way, with the sole use of a smartphone. In the research, smart irrigation, which aims to supply the plant’s water needs at the right time with the correct amount of water, showed savings in water use and greater water control by producers for their plantations, using a smartphone app ( Lopes et al. 2021 ). According to Varona & Zayas 2016 , correct irrigation of the coffee plantation at the correct period can lead to a decrease in production costs, as it contributes to the physiological need of the plant at that time. This opens up a financial margin for coffee farmers to invest in new techniques and technologies aimed at managing climate risk. In addition, water control in the plantation is essential for the Alta Mogiana region due to the reduction in rainfall observed in statistical tests, as it avoids wasting water.

However, according to Tran et al. 2021 , the expansion of irrigation in a given region must be accompanied by public policies for rural credit, in addition to professional qualification programs, for a better use of the technique. The need to qualify farmers is supported by Sasaki et al. 2013 in the use of electrostatic sprayers, which also requires skilled labor. Such sprayers presented savings in the use of agrochemicals in the plantation, contributing to lessening environmental damage ( Sasaki et al. 2013 ) and, consequently, to mitigating climate change.

Another promising alternative that can be integrated into coffee crops is nutritional control of the plant. For Ramírez-Builes & Küsters 2021 , nutrition with calcium and potassium showed a nutritive increase in the plant, causing it to develop greater resistance to biotic factors (such as pests and diseases) and abiotic factors (such as changes in precipitation and temperature) present in the environment. However, it is pointed out that, like smart irrigation and electrostatic spraying, plant nutrition also requires skilled workforce.

CONCLUSION

To analyze the exposure of coffee trees to climate risk in theAlta Mogiana region, one of the most important coffee-producingregions in Brazil, climate trend tests were carried out, which, ingeneral, suggest an increase in maximum and minimum temperaturesand a decrease in precipitation during the period 1991-2021. It ispossible to say that changes in temperature and precipitationpatterns have been occurring in the region within the past 30years.

Statistical tests showed a trend of increase in daily maximumtemperatures in the region, with the highest increase (1.4°C) in themunicipality of Pedregulho. The range of maximum temperatureincrease was 1.2°C to 1.4°C. Daily minimum temperatures also showedan increase, with the municipalities of Buritizal, Jeriquara,Ribeirão Corrente, Franca, Itirapuã, Patrocínio Paulista, andRestinga presenting the highest increase (0.8°C) in the region. Therange of minimum temperature increase was 0.6C to 0.8C. For therainfall statistical tests, the highest difference was observed inthe municipalities of Batatais, Nuporanga, and São José da BelaVista, where a daily average difference of 0.9mm was observed. Theother municipalities showed differences of 0.8mm. These valuessuggest that the annual rainfall deficit ranges between 292 and 328mm/rain per year. The maximum temperatures of coffee floweringmonths also presented an increasing trend, the highest increase(2.4°C) being in the municipality of Pedregulho. This may affect theflowering of the coffee trees and, thus, the quantity and qualityof the fruits.

Thus, one can conclude that there is greater exposure of coffeecrops to climate risk in the region, corroborating the IPCC reportson climate change already underway on the planet and its potentialimpacts on agricultural production. Thus, measures must be taken tomitigate losses in production related to the quantity of bagsproduced and the quality of the fruits.

Thus, from the literature review, it was possible to concludethat despite the greater exposure of coffee cultivation to climatechange in Alta Mogiana, possible solutions can be adopted asclimate risk management strategies. The literature review pointedto agroforestry systems and the use of irrigation in crops as themost prominent initiatives. Agroforestry systems contribute to thethermal protection of the plant against rising temperatures, inaddition to providing ecosystem services such as pollination andenvironmental preservation, helping mitigate climate change. Cropirrigation was a promising solution for places where changes in thepattern of precipitation distribution have been observed, asinferred from the statistical tests for the Alta Mogiana region.However, other initiatives, such as research on cultivars that aremore resistant to drought, pests, and diseases; training coursesfor farmers; increased credit and rural insurance; andelectrostatic spraying and nutritional control of the plant haveshown great potential for mitigating climate change.

ACKNOWLEDGMENTS

The authors thank Espaço da Escrita – Pró-Reitoria de Pesquisa –UNICAMP - for the language services provided. This work is supported bythe BI0S - Brazilian Institute of Data Science, grant #2020/09838-0,Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP). Theauthors thank the Conselho Nacional de Desenvolvimento Científico eTecnológico (CNPq), grant 403858/2021-6. This study was financed in partby the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior –Brazil (CAPES) – Finance Code 001.

• ABOELKHAIR H, MORSY M & EL AFANDI G. 2019. Assessment of Agroclimatology NASA POWER Reanalysis Datasets for Temperature Types and Relative Humidity at 2 m Against Ground Observations over Egypt. Adv Space Res 64(1): 129–142. doi: 10.1016/j.asr.2019.03.032 .
• ADGER WN. 2006. Vulnerability. Glob Environ Change 16(3): 268–281. doi: 10.1016/j.gloenvcha.2006.02.006.
• ALFONSI WMV, COLTRI PP, ZULLO JÚNIOR J, PATRÍCIO FRA, GONÇALVES RR DO V, SHINJI K, ALFONSI EL & KOGA-VICENTE A. 2019. Geographical Distribution of the Incubation Period of Coffee Leaf Rust in Climate Change Scenarios. Pesq Agropec Bras 54: 1–11. doi: 10.1590/s1678-3921.pab2019.v54.00273.
• ALVARENGA MIN & MARTINS M. 2004. Fatores Edáficos de Cafezais Arborizados. Arborização de Cafezais No Brasil. Vitória da Conquista: UESB.
• ALVARES CA, STAPE JL, SENTELHAS PC, MORAES GONÇALVES JL DE & SPAROVEK G. 2013. Köppen’s Climate Classification Map for Brazil. Meteorol Z 22(6): 711–728. doi: 10.1127/0941-2948/2013/0507.
• ASSAD ED ET AL. 2000. Zoneamento Climático Da Cultura Do Café (Coffea Arabica) Para o Sudoeste Do Estado Da Bahia. Comunicado Técnico 36: 1–6.
• ASSAD ED, PINTO HS, ZULLO JUNIOR J & ÁVILA AMH. 2004. Impacto Das Mudanças Climáticas No Zoneamento Agroclimático Do Café No Brasil. Pesq Agropec Bras 39(11): 1057–1064. doi: 10.1590/S0100-204X2004001100001.
• ASSAD E & PINTO H. 2008. Aquecimento Global e a Nova Geografia Da Produção Agrícola No Brasil. EMBRAPA-CEPAGRI.
• ASSUNÇÃO P & WANDER A. 2014. Competitividade Do Sistema Agroindustrial Do Feijão-Comum No Estado de Goiás. Scient Ple 10: 1–12.
• B MJ, R S, R GAW & DE ASR. 2010. Cultura de Café No Brasil: Manual de Recomendações. Varginha: Fundação PROCAFÉ.
• BAI J, CHEN X, DOBERMANN A, YANG H, CASSMAN KG & ZHANG F. 2010. Evaluation of NASA Satellite- and Model-Derived Weather Data for Simulation of Maize Yield Potential in China. Agro J 102(1): 9–16. doi: 10.2134/agronj2009.0085.
• CEDEPLAR & FIOCRUZ. 2008. Mudanças Climáticas, Migrações e Sáude: Cenários Para o Nordeste Brasileiro, 2000-2050. Belo Horizonte: Relatório de Pesquisa
• CUTTER SL & FINCH C. 2008. Temporal and Spatial Changes in Social Vulnerability to Natural Hazards. Proceedings of the National Academy of Sciences 105(7): 2301–2306. doi: 10.1073/pnas.0710375105.
• DE LEIJSTER V, SANTOS MJ, WASSEN MW, CAMARGO GARCÍA JC, LLORCA FERNANDEZ I, VERKUIL L, SCHEPER A, STEENHUIS M & VERWEIJ PA. 2021. Ecosystem Services Trajectories in Coffee Agroforestry in Colombia over 40 Years. Ecosys Serv 48: 1–13. doi: 10.1016/j.ecoser.2021.101246.
• DUBREUIL V, FANTE KP, PLANCHON O & SANT’ANNA NETO JL. 2019. Climate Change Evidence in Brazil from Köppen’s Climate Annual Types Frequency. Int J Climatol 39(3): 1446–1456. doi: 10.1002/joc.5893.
• DUFEK AS & AMBRIZZI T. 2008. Precipitation Variability in São Paulo State, Brazil. Theor Appl Climatol 93(3-4): 167–178. doi: 10.1007/s00704-007-0348-7.
• FALEIROS GD, BLISKA JÚNIOR A, NOGUEIRA TURCO PH & DE MELLO BLISKA FM. 2020. Avaliação Do Grau de Gestão Dos Cafezais Da Alta Mogiana Paulista. Cient Jabot 48: 1–16. doi: 10.15361/1984-5529.2020v48n1p01-16.
• FÜSSEL H-M. 2007. Adaptation Planning for Climate Change: Concepts, Assessment Approaches, and Key Lessons. Sust Sci 2(2): 265–275. doi: 10.1007/s11625-007-0032-y .
• GEORGE D, CLEWETT J, BIRCH C, WRIGHT A & ALLEN W. 2009. A Professional Development Climate Course for Sustainable Agriculture in Australia. Environ Educ Res 15(4): 417–441. doi: 10.1080/13504620902946978.
• GOES TB & CHINELATO GA. 2018. Viabilidade Econômico-Financeira Da Cultura Do Café Arábica Na Região Da Alta Mogiana. Rev IPecege 4(4): 31–39. doi: 10.22167/r.ipecege.2018.4.31.
• GONÇALVES N, ANDRADE D, BATISTA A, CULLEN L, SOUZA A, GOMES H & UEZU A. 2021. Potential Economic Impact of Carbon Sequestration in Coffee Agroforestry Systems. Agroforest Sys 95(2): 419–430. doi: 10.1007/s10457-020-00569-4.
• GORNALL J, BETTS R, BURKE E, CLARK R, CAMP J, WILLETT K & WILTSHIRE A. 2010. Implications of Climate Change for Agricultural Productivity in the Early Twenty-First Century. Phil Tran R Soc B 365(1554): 2973–2989. doi: 10.1098/rstb.2010.0158.
• HEINEMANN AB, RAMIREZ-VILLEGAS J, STONE LF & DIDONET AD. 2017. Climate Change Determined Drought Stress Profiles in Rainfed Common Bean Production Systems in Brazil. Agricul For Meteo 246: 64–77. doi:10.1016/j.agrformet.2017.06.005
• HERNANDES JL, PEDRO JÚNIOR MJ & BARDIN L. 2004. Variação Estacional Da Radiação Solar Em Ambiente Externo e No Interior de Floresta Semidecídua. Rev Árv 28(2): 167–172. doi: 10.1590/S0100-67622004000200002.
• HIRSCH RM & SLACK JR. 1984. A Nonparametric Trend Test for Seasonal Data With Serial Dependence. Water Resour Res 20(6): 727–732. doi: 10.1029/WR020i006p00727.
• HOOGENBOOM G. 2000. Contribution of Agrometeorology to the Simulation of Crop Production and Its Applications. Agric For Meteorol 103(1-2): 137–157. doi: 10.1016/S0168-1923(00)00108-8.
• JARAMILLO J, MUCHUGO E, VEGA FE, DAVIS A, BORGEMEISTER C & CHABI-OLAYE A. 2011. Some Like It Hot: The Influence and Implications of Climate Change on Coffee Berry Borer (Hypothenemus Hampei) and Coffee Production in East Africa. PLoS ONE 6: 1–14. doi: 10.1371/journal.pone.0024528.
• KARMESHU N. 2012. Trend Detection in Annual Temperature & Precipitation Using the Mann Kendall Test – A Case Study to Assess Climate Change on Select States in the Northeastern United States. Master of Environmental Studies Capstone Projects doi: 10.54302/mausam.v66i1.360.
• KATH J, BYRAREDDY VM, MUSHTAQ S, CRAPARO A & PORCEL M. 2021. Temperature and Rainfall Impacts on Robusta Coffee Bean Characteristics. Clim Risk Manag 32: 1–15. doi: 10.1016/j.crm.2021.100281.
• KENDALL MG. 1975. Rank Correlation Measures. London: Charles Griffin.
• KOBAYASHI K & SALAM MU. 2000. Comparing Simulated and Measured Values Using Mean Squared Deviation and Its Components. Agro J 92(2): 345–352. doi: 10.2134/agronj2000.922345x.
• KOH I, GARRETT R, JANETOS A & MUELLER ND. 2020. Climate Risks to Brazilian Coffee Production. Environ Res Lett 15(10): 1–13. doi: 10.1088/1748-9326/aba471.
• LIN BB. 2006. Agroforestry Management as an Adaptive Strategy Against Potential Microclimate Extremes in Coffee Agriculture. Agric For Meteorol 144(1-2): 85–94. doi: 10.1016/j.agrformet.2006.12.009.
• LOPES LCL, VIEIRA HD, VIEIRA GHS & SOUZA EF DE. 2021. Mobile Application Project for Conilon Coffee Irrigation Management. J Irrig Drain Eng 147(7): 04021021. doi: 10.1061/(ASCE)IR.1943-4774.0001574.
• LÓPEZ-SAMPSON A ET AL. 2020. Long-Term Effects of Shade and Input Levels on Coffee Yields in the Pacific Region of Nicaragua. Bois Et Forets Des Tropiques 346: 21–33. doi: 10.19182/bft2020.346.a36292.
• MANN HB. 1945. Nonparametric Tests Against Trend. Econometrica 13(3): 245. doi: 10.2307/1907187.
• MARENGO JA, JONES R, ALVES LM & VALVERDE MC. 2009. Future Change of Temperature and Precipitation Extremes in South America as Derived from the PRECIS Regional Climate Modeling System. Int J Climatol 30: 1–15. doi: 10.1002/joc.1863.
• MARENGO JA & CAMARGO CC. 2009. Surface Air Temperature Trends in Southern Brazil for 1960–2002. Int J Climatol 28(7): 893–904. doi: 10.1002/joc.1584.
• MARTINS FB, GONZAGA G, DOS SANTOS DF & REBOITA MS. 2018. CLASSIFICAÇÃO CLIMÁTICA DE KÖPPEN E DE THORNTHWAITE PARA MINAS GERAIS: CENÁRIO ATUAL E PROJEÇÕES FUTURAS. Rev Bras Climat 14: 129–156. doi: 10.5380/abclima.v1i0.60896.
• MARTINS MA, TOMASELLA J & DIAS CG. 2019. Maize Yield Under a Changing Climate in the Brazilian Northeast: Impacts and Adaptation. Agric Water Manag 216: 339–350. doi: 10.1016/j.agwat.2019.02.011.
• MENEZES JA, CONFALONIERI U, MADUREIRA AP, DUVAL I DE B, SANTOS RB DOS & MARGONARI C. 2018. Mapping Human Vulnerability to Climate Change in the Brazilian Amazon: The Construction of a Municipal Vulnerability Index. Plos One 13: 1–16. doi: 10.1371/journal.pone.0190808.
• MESQUITA CM, MELO EM DE, REZENDE JE DE, CARVALHO JS, FABRI JÚNIOR MA, MORAES NC, DIAS PT, CARVALHO RM DE & ARAÚJO WG DE. 2016. Manual Do Café: Implantação de Cafezais Coffea Arábica L. Belo Horizonte: EMATER-MG.
• MORETTIN PA & TOLOI CMC. 2006. Análise de Séries Temporais. São Paulo: Blucher.
• NEVES A. 2012. Detecção de Tendências No Padrão Temporal de Variáveis Hidrológicas. Aplicação à Precipitação a Diferentes Escalas Temporais. Master’s Thesis, University of Beira Interior, Covilha, Portugal (Unpublished).
• OBERMAIER M & ROSA LP. 2013. Mudança Climática e Adaptação No Brasil: Uma Análise Crítica. Estudos Avançados 27(78): 155–176. doi: 10.1590/S0103-40142013000200011.
• OJEDA JJ, VOLENEC JJ, BROUDER SM, CAVIGLIA OP & AGNUSDEI MG. 2017. Evaluation of Agricultural Production Systems Simulator as Yield Predictor of Panicum Virgatum and Miscanthus x Giganteus in Several US Environments. GCB Bioe 9(4): 796–816. doi: 10.1111/gcbb.12384.
• PETEK MR, SERA T & FONSECA IC DE B. 2009. Exigências Climáticas Para o Desenvolvimento e Maturação Dos Frutos de Cultivares de Coffea Arabica. Bragantia 68(1): 169–181. doi: 10.1590/S0006-87052009000100018.
• PETTITT AN. 1979. A Non-Parametric Approach to the Change-Point Problem. J R Stat Soc Ser C Appl Stat 28(2): 126. doi: 10.2307/2346729.
• PINHEIRO AG, SOUZA LSB DE, JARDIM AM DA RF, ARAÚJO JÚNIOR G DO N, ALVES CP, SOUZA CAA DE, SILVA GÍN DA & SILVA TGF DA. 2021. Importância Dos Modelos de Simulação de Culturas Diante Os Impactos Das Alterações Climáticas Sobre a Produção Agrícola - Revisão. Rev Bras Geo Fís 14(6): 3648–3666. doi: 10.26848/rbgf.v14.6.p3648-3666.
• PORFIRIO LL, NEWTH D, FINNIGAN JJ & CAI Y. 2018. Economic Shifts in Agricultural Production and Trade Due to Climate Change. Palgrave Commun 4(1): 111. doi: 10.1057/s41599-018-0164-y.
• QUINTÃO AF, BRITO I, OLIVEIRA F, MADUREIRA AP & CONFALONIERI U. 2017. Social, Environmental, and Health Vulnerability to Climate Change: The Case of the Municipalities of Minas Gerais, Brazil. J Environ Public Health 2017: 1–8. doi: 10.1155/2017/2821343.
• RAMÍREZ-BUILES VH & KÜSTERS J. 2021. Calcium and Potassium Nutrition Increases the Water Use Efficiency in Coffee: A Promising Strategy to Adapt to Climate Change. Hydrology 8(2): 75. doi: 10.3390/hydrology8020075.
• RESENDE NC, MIRANDA JH, COOKE R, CHU ML & CHOU SC. 2019. Impacts of Regional Climate Change on the Runoff and Root Water Uptake in Corn Crops in Parana, Brazil. Agric Water Manag 221: 556–565. doi: 10.1016/j.agwat.2019.05.018.
• RODRIGUES NA & REIS EA DOS. 2014. Influências Dos Fatores Climáticos No Custo de Produção Do Café Arábica. Custos e Agronegócio on-Line 10.
• SANTOS DF DOS, MARTINS FB, TORRES RR & UNIVERSIDADE FEDERAL DE ITAJUBÁ, BRAZIL. 2017. Impacts of Climate Projections on Water Balance and Implications on Olive Crop in Minas Gerais. Rev Bras Eng Agríc Ambient 21(2): 77–82. doi: 10.1590/1807-1929/agriambi.v21n2p77-82.
• SANTOS LF DOS, MARTINS FB & GARCIA SR. 2018. PADRÕES CLIMATOLÓGICOS DE PRECIPITAÇÃO E TEMPERATURA DO AR ASSOCIADOS AO RENDIMENTO DO FEIJÃO COMUM EM MINAS GERAIS. Rev Bras Climatol 1: 3–24. doi: 10.5380/abclima.v1i0.59108.
• SASAKI RS, TEIXEIRA MM, FERNANDES HC & MONTEIRO PM DE B. 2013. Deposição e Uniformidade de Distribuição Da Calda de Aplicação Em Plantas de Café Utilizando a Pulverização Eletrostática. Ciên Rur 43(9): 1605–1609. doi: 10.1590/S0103-84782013000900011.
• SAVARY S, NELSON A, WILLOCQUET L, PANGGA I & AUNARIO J. 2012. Modeling and Mapping Potential Epidemics of Rice Diseases Globally. Crop Prot 34: 6–17. doi: 10.1016/j.cropro.2011.11.009.
• SAYAGO S, OVANDO G, ALMOROX J & BOCCO M. 2020. Daily Solar Radiation from NASA-POWER Product: Assessing Its Accuracy Considering Atmospheric Transparency. Inter J Remote Sens 41(3): 897–910. doi: 10.1080/01431161.2019.1650986.
• SCHAEFFER RAS, SZKLO AS, LUCENA AFP DE, SOUZA RR DE, MOREIRA BS, BORBA C, COSTA IVL DA, PEREIRA JÚNIOR AO & CUNHA SHF DA. 2008. Mudanças Climáticas e Segurança Energética No Brasil. COPPE-UFRJ : 65.
• SHIFERAW B, TESFAYE K, KASSIE M, ABATE T, PRASANNA BM & MENKIR A. 2014. Managing Vulnerability to Drought and Enhancing Livelihood Resilience in Sub-Saharan Africa: Technological, Institutional and Policy Options. Weather Clim 3: 67–79. doi: 10.1016/j.wace.2014.04.004.
• SONALI P & NAGESH KUMAR D. 2013. Review of Trend Detection Methods and Their Application to Detect Temperature Changes in India. J Hydrol 476: 212–227. doi: 10.1016/j.jhydrol.2012.10.034.
• SOUZA DTM & HALAK AL. 2012. Agentes Polinizadores e Produção de Grãos Em Cultura de Café Arábica Cv. “Catuaí Vermelho.” Cient Jabot 40: 1–11.
• SPARKS A. 2018. Nasapower: A NASA POWER Global Meteorology, Surface Solar Energy and Climatology Data Client for R. J Open Source Software 3(30): 1–3. doi: 10.21105/joss.01035.
• TABARI H, MAROFI S, AEINI A, TALAEE PH & MOHAMMADI K. 2011. Trend Analysis of Reference Evapotranspiration in the Western Half of Iran. Agric For Meteorol 151(2): 128–136. doi: 10.1016/j.agrformet.2010.09.009.
• TANASIJEVIC L, TODOROVIC M, PEREIRA LS, PIZZIGALLI C & LIONELLO P. 2014. Impacts of Climate Change on Olive Crop Evapotranspiration and Irrigation Requirements in the Mediterranean Region. Agric Water Manag 144: 54–68. doi: 10.1016/j.agwat.2014.05.019.
• TAVARES P DA S, GIAROLLA A, CHOU SC, SILVA AJ DE P & LYRA A DE A. 2018. Climate Change Impact on the Potential Yield of Arabica Coffee in Southeast Brazil. Reg Envir Change 18(3): 873–883. doi:10.1007/s10113-017-1236-z
• THOMAZIELLO RA, FAZUOLI LC, PEZZOPANE JR, FAHL JI & CARELLI MLC. 2000. Café Arábica: Cultura e Técnicas de Produção. Campinas: Instituto Agronômico.
• TORRES GAL, PANTANO AP & CAMPAROTTO LB. 2020. Variabilidade Pluviométrica de Franca e Mococa Regiões Produtoras de Café No Estado de São Paulo / Variabilidade Pluviométrica de Franca e Mococa Regiões Produtoras de Café No Estado de São Paulo. Braz J of Anim and Environ Res 3(4): 2829–2836. doi: 10.34188/bjaerv3n4-005.
• TRAN DNL, NGUYEN TD, PHAM TT, RAÑOLA RF & NGUYEN TA. 2021. Improving Irrigation Water Use Efficiency of Robusta Coffee (Coffea Canephora) Production in Lam Dong Province, Vietnam. Sustainability 13(12): 6603. doi: 10.3390/su13126603.
• VAN WART J, GRASSINI P, YANG H, CLAESSENS L, JARVIS A & CASSMAN KG. 2015. Creating Long-Term Weather Data from Thin Air for Crop Simulation Modeling. Agric Forest Meteorol 209-210: 49–58. doi: 10.1016/j.agrformet.2015.02.020.
• VARONA CRM & ZAYAS CEC. 2016. Viabilidad Económica Del Riego Localizado En El Cultivo Del Cafeto. Viabilidad Económica Del Riego Localizado En El Cultivo Del Cafeto 25: 44–50.
• WMO- WORLD METEOROLOGICAL ORGANIZATION. 2019. Standards and recommended practices. URL https://public.wmo.int/en/resources/standardstechnical-regulations Accessed January 2022.
  » https://public.wmo.int/en/resources/standardstechnical-regulations
• XAVIER AC, KING CW & SCANLON BR. 2016. Daily Gridded Meteorological Variables in Brazil (1980-2013). Inter J Climatol 36(6): 2644–2659. doi: 10.1002/joc.4518.
• YAGHOUBI F, BANNAYAN M & ASADI G-A. 2020. Performance of Predicted Evapotranspiration and Yield of Rainfed Wheat in the Northeast Iran Using Gridded Ag MERRA Weather Data. Inter J Biometeo 64(9): 1519–1537. doi: 10.1007/s00484-020-01931-y.
• ZHANG S & LU XX. 2009. Hydrological Responses to Precipitation Variation and Diverse Human Activities in a Mountainous Tributary of the Lower Xijiang, China. CATENA 77(2): 130–142. doi: 10.1016/j.catena.2008.09.001.

Publication Dates

• Publication in this collection
  21 Oct 2022
• Date of issue
  2022

History

• Received
  15 Oct 2021
• Accepted
  4 May 2022