This study analyzed the temporal and spatial patterns of the precipitation (PRP) and maximum temperature (TMAX) time series of 37 municipalities proportionally distributed in the northeastern part of Pará (Eastern Amazon), from 1981 to 2018 (37 years). Environmental data on citrus productivity-PROD (oranges and lime) (IBGE), evapotranspiration (ET) and deforestation (PRODES) were used for the last 15 years to identify the impacts of climate change. The non-parametric approach, specifically the Mann-Kendall (MK) test and delta variation in the period 1989-2003 and 2004-2018 (15 years), were applied to estimate annual and seasonal trends and detect variations after using heatmap and boxplots. To compose the climate change study, Principal Component Analysis (PCA) was applied to environmental variables and Principal Component Regression (PCR) was used to test the hypothesis. The results showed an increasing positive trend in the annual and seasonal TMAX time series over 37 years, with the most significant trends in the dry season. The trend results in PRP, on the other hand, showed regions with both negative and positive trend significance. On annual and seasonal scales, regions such as R5 in northeastern Pará showed positive trends. On the other hand, negative trends of decreasing PRP were observed in the coastal areas of Pará (R1 and R3). Based on PCA and PCR results, deforestation data are associated with higher maximum temperatures and orange productivity follows correlated with PRP and ET. In contrast, PRP and ET are considered limiting factors for lime productivity.