Plant sampling uncertainty: a critical review based on moss studies

Abstract

Estimates of the distribution, migration, and accumulation of trace elements using mosses as bioindicators have successfully been used in biomonitoring studies since at least the 1970s. Chemical analysis of moss samples is also an important tool for assessing concentrations of elements in analyzed material at a given time. To achieve satisfactory accuracy in environmental studies, the best sampling approach must be used. Methods for the estimation of uncertainty derived from analytical procedures are well recognized, but the errors generated as a result of sampling are very often overlooked. Sampling uncertainty can be managed by a judicious selection of the sampling method, the amount of samples collected, and by following appropriate type of sampling protocols. The sampling protocol generally contains information about location of sampling sites, time of sampling (e.g., season), the species collected, type of sample (single, sub-sample), and monitored parameters (e.g., climate, analyzed substances). Information about seasonal variability; topographic, climatic, edaphic, and hydrologic conditions (type and amount of precipitation, rosewind); age; and part of plant that was collected is often ignored. There is no precise information on how these factors affect the sampling step and overall uncertainty over what procedures must be followed to reduce errors derived from plant sampling. This information is necessary when long-range and comparative studies are conducted. In this paper, we review how individual factors, such as (i) type of sampling strategy, (ii) representative sampling, (iii) seasonal variability, and (iv) which part of the plant is collected, may influence the concentration of trace elements in moss tissues and the level of uncertainty associated with sampling. In addition, we also discuss plant sample preparation techniques and how this may cause an uncontrolled element loss.