Tropical planted forests have demonstrated that tree plantations can have variable growth and water use patterns in response to drought. Yet research on how specific species will perform during a drought and whether this response can be mediated through forest management is still poorly understood. However, this information is vital as planted forests become more common throughout the tropics, where more extreme droughts are forecasted. We took advantage of a historic El Niño driven drought in 2015-2016 in central Panama to test how native species and a non-native species responded to reduced rainfall (60% less precipitation than average). We installed 96 trees with sap flow sensors that measured continuously for 2.5 years. We also measured meteorological data during the same period (vapor pressure deficit [VPD], temperature, relative humidity, precipitation) as well as the growth of the planted trees, to better understand what factors contributed most to changes in tree water use. In the non-native (Tectona grandis) plantation, we also conducted a thinning experiment, thinning half of the plots by 50% to see if thinning may mitigate the effects of drought. While one of the native species (Dalbergia retusa) we studied maintained daily transpiration before and during the drought, another native species (Terminalia amazonia) and non-native species T. grandis showed decreases in daily transpiration during the drought, but significantly higher daily transpiration rates than D. retusa. Additionally, we found that thinning had short-term effects on plant water use and growth, with initial increases in water use in thinned stands (versus unthinned stands) during the drought, but that once the drought ended, water use returned back to similar rates as the unthinned treatment. Notably, we found that each species had a unique temperature and VPD threshold, where the sap flow began to decline. While VPD generally led to increases in sap flow, this declined for D. retusa at 32°C and >34°C and 28°C for T. amazonia and T. grandis, respectively. Our results find that D. retusa is particularly water-use efficient compared to the other two species and may be an important reforestation species because of this trait. Notably, we also found differential temperature thresholds, where at high VPDs, sap flow began to decline. Of all three species, T. grandis had the lowest temperature threshold, which could pose a problem in future climate scenarios, where temperatures are expected to increase. Our results have important implications for reforestation efforts, where accurate tree species selection is critical.