Pulp and paper technical guidance: criteria and guidance for pronounced eutrophication, chapter 4

Eutrophication Criteria

The general criteria for eutrophication usually found in the published scientific literature tend to reflect the potential risks to the aquatic environment and do not necessarily predict the presence of negative impacts. Usually, common water quality indicators are used to define the degree of eutrophication. The most commonly used are: total phosphorus, chlorophyll a and Secchi disc depth (Gray et al., 2002).

Published scientific literature based on benthic studies indicates that the level of eutrophication can be reflected by variations in benthic invertebrate abundance (number of animals per unit area, or density) and taxon richness (number of distinct family level taxa per sample) (Grall and Chauvaux, 2002; Nixon, 1995). Such endpoints have been calculated for most of the sites during the EEM program and are readily available.

Figure 1, based on Lowell et al. (2003), summarizes gradual variations in abundance and taxon richness along a gradient of nutrient enrichment, toxicity or smothering and generalizes the particular benthic response patterns stated in the Pulp and Paper Guidance Document (Chapter 12, Figure 12-2 in Environment Canada, 2005a). Based on this widely accepted response pattern, pronounced eutrophication is characterized in benthic invertebrate communities by elevated abundance of individuals from so-called opportunistic species (e.g. freshwater tubificid worms and chironomids) and localized extirpation of sensitive fauna, normally larger taxa (e.g. mollusks). Hyper eutrophication, on the other hand, is usually characterized by a decline in the abundance or biomass of benthic organisms; heavy sedimentation of organic matter can smother zoobenthos making the habitat available for opportunistic colonizers (e.g. polychaetes) (Gray et al., 2002).

Figure 1. Variations in benthic invertebrate abundance (density) and taxon richness in an exposure site relative to a reference site along a gradient of increasing nutrient enrichment, producing either a toxicity or a smothering effect.
This graph is based on a figure from Lowell et al., 2003.
Note: the lines illustrating + and - 2SD for abundance and taxon richness are theoretical. For the significance of cases 1, 2 and 3: see text.
*SD - standard deviations

For the purposes of EEM, based on scientific information collected during the program and analysis of Figure 1, the following criteria have been developed and are proposed for use in defining pronounced and hyper eutrophication (Table 1).

Table 1. Definitions of pronounced and hyper-eutrophication applied to benthic invertebrate community EEM data*.

Notes:

• + infers a statistically significant increase of at least 2 standard deviations relative to mean of the reference area.
• - infers a statistically significant decrease of at least 2 standard deviations relative to mean of the reference area.
• * applicable for control-impact, artificial substrate and mesocosm studies. For gradient designs, the pronounced eutrophication criterion is a statistically significant decrease in abundance and/or increase in taxon richness with increasing distance from the outfall (i.e., equivalent to increased abundance and/or decreased richness in more effluent-exposed areas close to the outfall). The hyper-eutrophication criterion is a statistically significant increase in abundance and taxon richness with increasing distance from the outfall (i.e., equivalent to decreased abundance and richness in more effluent-exposed areas close to the outfall).

Data from EEM studies showed that most mills reported statistically significant effects in at least one of the core endpoints for benthic invertebrate and fish population studies. Since it was recognized that statistically significant differences are not necessarily considered to be serious, a concept of Critical Effect Sizes (CES) was developed to identify differences that could lead to ecologically significant effects. CES for benthic invertebrate community endpoints are based on whether the change in the endpoints falls out of the natural range of variability using a statistical basis of two standard deviations (?2 SD). The same concept has been retained to develop the eutrophication criteria: 1) statistical significance and 2) an effect size exceeding two standard deviations.

Many mills use gradient designs rather than control/impact designs and, for those mills, reported increases or decreases are based on statistically significant correlations (value of endpoint versus distance of station from outfall). Some mills have opted for benthic studies based on artificial substrates or mesocosm studies. For those kinds of studies the same CES were applied.

Based on Figure 1 and Table 1, benthic survey data defines pronounced eutrophication when mean exposure abundance (density) is significantly greater than the reference abundance mean, and the difference is greater than +2 SD of the reference mean (Case number 1 in Figure 1).

Another scenario indicative of pronounced eutrophication is defined when the mean exposure taxon richness is significantly lower than the mean reference taxon richness, and the reduction exceeds 2 SD of the reference mean (Case number 2 in Figure 1).

A scenario reflective of conditions worse than pronounced eutrophication, called hyper-eutrophication (Table 1) has been defined in cases where both exposure abundance and taxon richness means are significantly smaller than reference mean values and the difference exceeds 2 SD of the corresponding reference value (Case 3 in Figure 1).

Hyper-eutrophic conditions can occur in cases where an extensive loading of organic matter is combined with low dissolved oxygen levels. Such a situation can also take place during extensive particle deposition, thus “smothering” benthic habitats. Alternatively, it is possible that sediment-associated toxicity generates a response pattern for the benthos in the exposure area that is not attributable to nutrient loading but comparable to case number 3 of Figure 1 (“Toxicity or smothering”). However, each of these scenarios requires additional information to confirm which mechanism is the principal cause of effects in question. Regardless of the cause, mills falling into this latter category (Table 1) are showing serious effects and deserve further studies.