Subsampling and Sorting Protocol

Sorting macroinvertebrates from organic and inorganic material in the field is time consuming and inaccurate. Therefore, we recommend preservation of all collected material (T.F. Cuffney et al. 1993; Moulton et a. 2000). NAMC technicians seperate out invertebrates from samples using a microscope and place them in vials by order for taxonomists to identify (Merrit and Cummins 2008). It is cost prohibitive to identify all individuals in samples. Therefore, technicians subsample material until they have sorted out ~600 macroinvertebrates from each sample following Vinson and Hawkins (1996).

Benthic macroinvertebrate sorting using the Caton tray method can provide technicians with a quantitatively targeted approach to subsampling. In general, this method should enable a technician to avoid over-sampling and therefore can reduce total sample processing time. The approach is fundamentally like the sieve technique: total sample material is divided into spatially identical cells. The contents of a randomly selected cell are removed and counted. This process is repeated until target minimum counts are reached. Spaces can be fully removed, or partially removed using a tubular subsampler tool.  In all cases, it is imperative to accurately document total area (quantity of sample) processed by technician.

Guiding Principles

1. Achieve count as close to or > 600 as possible
2. Always remove material from at least 3 grid spaces (avoids statistical errors associated with clumping).
3. Start with full grid processing. Transition to tubular subsampler tool as needed to achieve final count slightly larger more closely than 600.

Specific Steps

1. Assess whether sample needs elutriation or other pre-splitting steps:
   
   1. Elutriate, discard rocks.
   2. Rinse off then discard large sticks and large leaves.
   3. Wash any sample greater than or equal to 4 jars through a sieve to reduce the amount of material to process. (also elutriate, wash leaves, etc.)
   4. Transfer contents of sample to Caton tray and uniformly distribute material.
2. Choose the appropriate grid dimensions for sample (e.g., 3x3 or 5x5 grid) based on the amount of material in sample.
3. Place the appropriate grid in the Caton tray on top of the sample, being careful to not crush any large organisms.
   a. If visible organisms fall between grid cells, move them into the grid cell containing the head. Once the organisms are in a cell, push down on the grid to crush any organic matter and create separation between cells.
4. Using random.org as a random number generator, select 3 numbers, which will correspond to the grid cells in the grid. 3 cells are the minimum required subsample, regardless of grid dimensions. Note these numbers in your notebook.
5. Remove the lowest numbered corresponding grid cell’s contents using a combination of spoons, forks, and an aspirator. Larger material should be removed using a utensil, whereas smaller alcohol-suspended material should be removed using the aspirator. Be sure to add more alcohol to a cell if the contents cannot be aspirated.  Work through the first cell and sort normally. Follow the dichotomous key below as grid cells are sorted to determine the appropriate workflow. After each cell/tubular subsample is completed, log the corresponding information in your notebook. For example:
   
   1. Random grid cell numbers: 2,5,19
      
      1. 2/15/2023: cell #2, 4%, 150 bugs, cell#5, 4%, 350 bugs, cell #19, 0.5%, 100 bugs.
      2. 2/15/2023: total split % = 8.5%. total bug count = 600, total hours = 4hrs.

Determining Subsampling Using the Caton Tray After Selecting Grid Size

Methodology

General Method - Starting a Sample

1. Pour the contents of the sample into the Caton tray after draining the alcohol into a waste container through a 500-micron sieve. Add just enough alcohol to moisten the contents of the tray, then physically move the tray to even distribute the contents.
2. Use your best judgement to determine if a 3x3 or a 5x5 grid is appropriate for your sample. E.g., if your sample is densely filled with organic matter, use the 5x5 grid.
3. After choosing the first three grid cells via random.org, use a scalpel to cut the edges of the first grid cell’s contents, allowing easy removal. Use a spoon or another appropriate utensil to remove the entire cell’s contents. Any leftover alcohol in the cell can be removed using the aspirator.
4. If using the tubular subsampler, drop the subsampler into the middle of the chosen cell. Press down on the subsampler to cut through the organic material using the beveled edge. If needed, use the scalpel to cut around the subsampler. Remove the contents of the subsampler using a spoon beneath the subsampler. Although there is no way to prove random distribution of organisms in each grid cell, the assumption is that all organisms are randomly distributed. The subsampler equates to ~0.5% of the sample. Use the Companion excel file to determine final split percent.
5. Although the protocol does have a distinct dichotomy between tubular subsampler or not, one may use a tubular subsampler if getting close to 600 bugs after finishing any number of total grid cells.
6. Use the key on the page above OR the flowchart below to walk through a sample.

Literature Cited

Cuffney, T.F. et al., 1993. Methods for collecting benthic invertebrate samples as part of the National Water-Quality Assessment Program, US Geological Survey, http://water.usgs.gov/nawqa/protocols/OFR-93-406/.

Merritt, R.W., Cummins, K.W. & Berg, M.B., 2008. An Introduction to the Aquatic Insects of North America 4th ed., Kendall Hunt Publishing.

Moulton, S. et al., 2000. Methods of analysis by the US Geological Survey National Water Quality Laboratory-processing, taxonomy, and quality control of benthic macroinvertebrate samples, US Geological Survey, http://nwql.usgs.gov/OFR-00-212.shtml.

Vinson, M.R. & Hawkins, C.P., 1996. Effects of sampling area and subsampling procedure on comparisons of taxa richness among streams. Journal of the North American Benthological Society, 15(3), p.392–399.