Temporal considerations for an effective sampling of personal protective equipment litter derived from the COVID-19 pandemic

Abstract

Personal protective equipment (PPE) has become a new pollutant derived from the COVID-19 pandemic. Much of the efforts to characterize PPE litter has focused on its spatial distribution (i.e., trying to identify hotspots of PPE litter), however, such efforts have been limited in the temporal domain, which might result in under- or overestimations in annual projections. Here, using 55 continuous days of sampling in an urban and tropical neighborhood in south east Mexico, I show that in order to have a robust and defensible average and variance values it is needed at least 22 days of random sampling. Nonetheless, this minimum number might change in different ecosystems and land use areas of the built environment due to the temporal variability of the human behavior and activities related to the surveyed areas, as well as the influence of weather conditions that might affect the mobility of people. Furthermore, I discuss how it is recommended to report the daily average density of PPE litter (items m−2 day−1) and its variability (i.e., 95 % confidence intervals), rather than only the density of PPE litter (items m−2) in order to facilitate annual estimates of PPE litter disposal.