Environmental monitoring using IoT sensors for ecosystem conservation

Objective: To evaluate water quality through a network of IoT sensors installed in conserved, transitional, and intervened areas of a riparian ecosystem, considering the variability of environmental parameters, the accuracy of the devices, and the operational performance of the system. Methodology: A quantitative, applied, non-experimental and longitudinal research was developed. Nine IoT stations were installed, distributed at three points in each environmental zone. For 90 days, temperature, pH, turbidity, electrical conductivity and dissolved oxygen were recorded every 30 minutes. Validation was carried out using 117 paired measurements with reference equipment. The data were analyzed using descriptive statistics, mixed linear models and coefficients of agreement. Results: 95.0% of the 38,880 scheduled registrations were recovered, with an operational availability of more than 94%. The intervened area presented higher temperature, turbidity and conductivity, as well as lower dissolved oxygen, with significant differences between the three zones. The sensors showed correlations between 0.93 and 0.98 and concordance coefficients greater than 0.90. Conclusions: The IoT network allowed the identification of spatial and temporal variations in water quality and generated measurements consistent with the reference equipment. Its application can complement conventional monitoring, detect critical conditions and support conservation strategies in riparian ecosystems.