Climate change significantly threatens agricultural productivity, with coffee being a highly vulnerable crop. Drought is a major environmental constraint, reducing yields by up to 80% in some regions. *Coffea canephora* (conilon coffee), a significant crop in Espírito Santo, Brazil, which accounts for 70% of Brazil's conilon production and approximately 15% of global production, is particularly susceptible to climate change impacts. Projections suggest that substantial areas currently suitable for conilon cultivation may become unsuitable due to rising temperatures and altered precipitation patterns. This is economically alarming because conilon coffee is responsible for a significant portion of Espírito Santo's GDP and provides many jobs. Previous research has shown the negative effects of water deficit on coffee growth, photosynthesis, and carbon metabolism, even in drought-tolerant genotypes. Drought years in Espírito Santo are associated with low rainfall during the winter and fall, coinciding with the harvest season. High temperatures exacerbate the negative impacts of drought, impairing plant metabolic processes and reducing yield. Agroforestry systems have shown potential to mitigate the negative effects of drought and high temperatures by improving the microclimate, soil moisture, and water infiltration. Espírito Santo experienced a record harvest in 2013/2014, followed by reduced yields in subsequent seasons, highlighting the vulnerability of the region to climate change. This study aims to analyze the impact of drought and high temperatures on conilon coffee production in Espírito Santo, identifying the most sensitive periods of the crop cycle and suggesting adaptation strategies.

Several studies have investigated the effects of water deficit on *Coffea canephora*. Pinheiro et al. (2004) found that net carbon assimilation rate decreased under drought stress, regardless of genotype. Praxedes et al. (2006) showed that drought led to decreased stomatal conductance and photosynthetic capacity. While drought-tolerant genotypes can increase water use efficiency, this is often associated with increased leaf temperatures, negatively impacting performance in warmer climates. DaMatta and Ramalho (2006) highlighted drought and extreme temperatures as major climatic limitations. Studies have shown that supra-optimal temperatures combined with water deficit intensify oxidative stress, leading to cell damage and leaf abscission. Agroforestry systems have been shown to mitigate microclimate variability, increase soil moisture, and reduce air temperatures, enhancing coffee resilience to climate change (Gidey et al., 2020). However, information on conilon coffee responses to drought and high temperature at a regional scale in Espírito Santo was lacking, which this study aimed to address.

This study encompassed nine municipalities in the north, northeast, and northwest regions of Espírito Santo, Brazil, major conilon coffee-producing areas. Data were collected from various sources: IBGE (production, yield, planted area), CONAB (production, yield, planted area at state/national level), Agritempo (monthly rainfall and air temperature), and Embrapa MODIS (Enhanced Vegetation Index – EVI). Data processing involved rescaling EVI values (0-1), calculating average EVI for three periods (P1: September-December; P2: January-March; P3: April-August), and statistical analysis. An identity test was used to compare the annual variation in production and planted area. Scatter plots analyzed the relationship between rainfall, temperature, and yield. Response surface methodology assessed the combined effects of rainfall and temperature on production. Principal component analysis identified the most influential rainfall period on yield. Finally, EVI maps were generated to spatially and temporally assess drought impacts.

The 2015/2016 season was the most severely affected by drought and high temperatures, with annual rainfall 40% below the average and a 41% reduction in production. The average annual air temperature during the two lowest yield seasons (2012/2013 and 2015/2016) was approximately 1°C higher than in preceding seasons. The identity test rejected the hypothesis that production reduction was solely due to a decrease in planted area. Increased air temperature showed a more significant negative impact on production than decreased annual precipitation. Scatter plots indicated a positive, though not always statistically significant, relationship between accumulated rainfall and yield across all nine municipalities. Response surface methodology revealed a significant interaction between temperature and rainfall, with lower yields observed at higher temperatures and lower rainfall across all periods (P1, P2, P3, annual). Principal component analysis indicated that rainfall during September-December (P1) and April-August (P3) most strongly influenced yield, while January-March (P2) had less influence. EVI maps showed lower values during 2014/2015 and 2015/2016, especially in P1 and P2, indicating drought stress.

The findings demonstrate the vulnerability of conilon coffee plantations in Espírito Santo to combined drought and high-temperature stress. The stronger negative impact of temperature compared to rainfall highlights the importance of considering temperature increases in future climate change predictions. The results align with previous studies showing the negative impacts of drought and heat stress on coffee growth and yield. The use of EVI effectively detected drought conditions at the regional scale, complementing the analysis of meteorological data. The identified critical periods in the coffee cycle (P1 and P3) can inform targeted irrigation strategies and other management practices. The study’s large-scale analysis offers a significant contribution to regional drought risk assessment. The limited statistical significance in some relationships highlights the complexity of factors influencing coffee yield at the municipal level.

This study integrated diverse data sources to assess the impact of drought and high temperatures on conilon coffee production in Espírito Santo. The 2015-2016 season was critically affected, revealing the vulnerability of the region to climate extremes. Increased temperature negatively impacted production more than decreased precipitation. Agroforestry systems and small reservoirs offer promising adaptation strategies. Future research should focus on controlled environment studies to better understand the effects of specific water stress levels and genotypes on coffee performance.

The study's spatial resolution is limited by the availability of data. The use of average values across municipalities may mask within-municipality variation in rainfall and other factors. While the study considered the effect of irrigation, it did not fully account for variations in irrigation practices among farms. The analysis did not directly measure physiological responses to drought and heat, relying on correlations between climatic variables, EVI, and yield.