Introduction.

Climate change adaptation measures require informed decision-making. However, small-scale agriculture shows low rates of technology adoption due to cost and lack of connectivity. Objective. To evaluate three low-cost prototypes of small-scale agriculture devices for soil moisture measurement in different soil textures, as well as to determine the respective calibration equations, and the effects of electrical conductivity and temperature on moisture measurement. Materials and methods. Three measurement and recording prototypes for soil moisture were evaluated in soils with variations in clay/sand content and electrical conductivity in productive plots at Zamorano University in Honduras and a demonstration farm in Popayán, Colombia during the first quarter of 2022. Commercial sensors were used as a reference to compare the performance of the prototypes through regression analysis of hourly sensor readings over 90 days. Electrical conductivity (dS/m) and temperature variables were collected to determine their influence on moisture reading accuracy. Results. The soil moisture measurement devices showed better performance in soils with lower sand content. The measurement from the devices overestimated moisture readings by 0.19 to 0.52 percentage points for each additional degree of soil temperature. Additionally, for each additional dS/m of electrical conductivity, the reading needed adjustment by 8 to 55 percentage points. Conclusions. Prototype A was the most accurate device, while prototype B was the most precise compared to commercial sensors. Soil moisture devices performed best in soils with lower sand content. The three models evaluated performed best in loam soil with a medium clay content.