Quantifying particulate matter deposition in Niwot Ridge, Colorado: Collection of dry deposition using marble inserts and particle imaging using the FlowCAM

Atmospheric wet and dry deposition are important sources of carbon for remote alpine lakes and soils. The carbon inputs from dry deposition in alpine National Atmospheric Deposition Program (NADP) collectors, including aeolian dust and biological material, are not well constrained due to difficulties in retaining particulate matter in the collectors. Here, we developed and tested a marble insert for dry deposition collection at the Niwot Ridge Long Term Ecological Research Station (NWT LTER) Soddie site (3345 m) between 24 May and 8 November 2011. We conducted laboratory tests of the insertʹs effect on particulate matter (PM) mass and non-purgeable organic carbon (DOC) and found that the insert did not significantly change either measurement. Thus, the insert may enable dry deposition collection of PM and DOC at NADP sites. We then developed a method for enumerating the collected wet and dry deposition with the Flow Cytometer and Microscope (FlowCAM), a dynamic-image particle analysis tool. The FlowCAM has the potential to establish morphology, which affects particle settling and retention, through particle diameter and aspect ratio. Particle images were used to track the abundance of pollen grains over time. Qualitative image examination revealed that most particles were biological in nature, such as intact algal cells and pollen. Dry deposition loading to the Soddie site as determined by FlowCAM measurements was highly variable, ranging from 100 to >230 g ha−1 d−1 in June–August 2011 and peaking in late June. No significant difference in diameter or aspect ratio was found between wet and dry deposition, suggesting fundamental similarities between those deposition types. Although FlowCAM statistics and identification of particle types proved insightful, our total-particle enumeration method had a high variance and underestimated the total number of particles when compared to imaging of relatively large volumes (60–125 mL) from a single sample. We recommend use of the FlowCAM, especially for subclasses of particles, but in light of uncertainty in particle counts, believe that it should be paired with traditional methods such as microscopy in this stage of the techniqueʹs development. Analysis of well-mixed samples produced lower variability than settling methods used for algae samples. Use of the marble inserts in the dry deposition collector in the NADP context is recommended, and the implications of various particle counting and identification methods are explored.