Dynamics of pesticides in surface water bodies by applying data mining to spatiotemporal big data. A case study for the Puglia Region

Abstract

Surface water pollution by pesticides is a primary concern in many parts of the world. Therefore, an effective monitoring program is essential to assess the environmental state of aquatic ecosystems and evaluate mitigation strategies.In the Puglia Region (southern Italy), a complex seasonal monitoring program for detecting pesticide residues in water was initiated in 2018 and is still underway (over four years). The program was based on site-specific assessments and identified 170 Plant Protection Products (PPPs) to identify residues in the surface water bodies of Puglia.In this context, the present work aims to analyse pesticide data obtained from the regional monitoring of rivers over four years using a multidisciplinary approach, which includes statistical methods, data mining, and mapping tools to extract as much information as possible.To this end, data mining was applied to identify the prevalent mixtures of pesticide residues found at the monitoring stations and correlate these results with possible causative factors. The results showed that surface water bodies are subject to different pressures derived from the massive use of PPPs, and several seasonal and territorial-related factors were identified to be strictly correlated with pesticide concentration results. Nine main PPPs mixtures have been identified in the Puglia River. Glyphosate, AMPA, imidacloprid, and azoxystrobin represent the main residues detected in the surface aquatic environments regarding the amount and frequency revealed.The methodological approach proposed in the present work can represent a good “model study” to be used by researchers to interpret water quality trends and data variability from long-term monitoring studies. Moreover, our results can significantly support the decision-making process and implementation of environmental mitigation measures by optimising the results of complex monitoring programs.