Biological Monitoring Covariates

When monitoring environmental impacts of anthropogenic activities, it is useful to collect ancillary data to use as covariates. These variables can help account for natural variation in the communities studies, which helps prevent their confounding of observations. In our work with benthic macroinvertebrate assemblages, we always collect flow measures for this purpose (near-substrate flow measures can account for much of the variation in simple benthic communities (e.g., Hart & Finelli 1999). With a sufficiently rigorous sampling design, the effects of flow on the community can be “teased” out of the analysis, so that the effects of anthropogenic stressors on benthic communities changes can be more accurately assessed–assuming, of course, that flow is not part of the anthropogenic impact in the study area.

Moss, macrophytes, and filamentous algae can also alter the abundance of certain types of macroinvertebrates. Moss, is not palatable to most macroinvertebrates because of the presence of protective chemicals. However, it does offer refuge from the sheer-forces of fast water. It also traps detritus (food), and can increase the amount of surface area available for colonization. Thus the presence of moss can have a strong influence on the structure of macroinvertebrate assembleges sampled in an assessment. When we want to account for this variation, we have found that the simplest (and effective) way, is to take a known amount of alcohol from the preserved samples and rank their Greenness. The Rank can be a very useful covariate when the data are analyzed. Other important covariates usually include a particle index, velocity, and depth.

The assumption is that the “Greenness” of the sample’s preservative is proportional to the amount of material soaking in alcohol for a given amount of time.  Thus, more greenness indicates there is more living plant material collected in the sample. Samples of similar greenness were probably influenced by living plants similarly, whereas the community composition of a very pale sample was probably influenced less by plant material than a sample with deep green preservative.

There are other nuances as well. For example, periphyton may cause a different kind of green tint to preservative than moss does–and it is likely to have a different influence on the benthic assemblage. Fortunately, for our purposes, moss seems to have much stronger effect on the sample’s color than periphyton. Also, you need to ensure that all samples are preserved with the same type and strength of preservative (95% ethanol, or 90% Isopropanol, or 90% denatured ethanol, not some of all three). Samples preserved in formalin would need to be transferred to alcohol before analysis–and even then should not be compared with samples preserved only with ethanol because of formalin’s ability to “fix” pigments.

Tags: , , , , , , , , , ,

Comments are closed.