Bioindication and modelling of atmospheric deposition in forests enable exposure and effect monitoring at high spatial density across scales

Context: For enhancing the spatial resolution of measuring
and mapping atmospheric deposition by technical devices and
by modelling, moss is used complementarily as bio-monitor.
Aims: This paper investigated whether nitrogen and heavy
metal concentrations derived by biomonitoring of atmospheric
deposition are statistically meaningful in terms of compliance with minimum sample size across several spatial levels (objective 1), whether this is also true in terms of geostatistical criteria such as spatial auto-correlation and, by this, estimated values for unsampled locations (objective 2) and whether moss indicates atmospheric deposition in a similar way as modelled deposition, tree foliage and natural surface soil at the European and country level, and whether they indicate site-specific variance due to canopy drip (objective 3).
Methods: Data from modelling and biomonitoring atmospheric
deposition were statistically analysed by means of
minimum sample size calculation, by geostatistics as well as
by bivariate correlation analyses and by multivariate correlation analyses using the Classification and Regression Tree approach and the Random Forests method.
Results: It was found that the compliance of measurements
with the minimum sample size varies by spatial scale and
element measured. For unsampled locations, estimation could
be derived. Statistically significant correlations between concentrations of heavy metals and nitrogen in moss and modelled atmospheric deposition, and concentrations in
leaves, needles and soil were found. Significant influence of canopy drip on nitrogen concentration in moss was proven.
Conclusion: Moss surveys should complement modelled atmospheric deposition data as well as other biomonitoring approaches and offer a great potential for various terrestrial monitoring programmes dealing with exposure and effects.