David R. Barton and Christine M. Boston
Department of Biology
University of Waterloo
Such simple indices can be very useful, but the amount of information revealed by the benthic fauna increases with sampling effort and closer identification of the animals, so the design of a benthic invertebrate monitoring program depends on the kinds of changes which need to be detected. Discrimination between streams which should be capable of supporting brook trout and those which clearly are not, is simple and inexpensive; assessment of the effects of urban development and any remedial efforts requir ed of developers demands greater effort. Decisions about the final design of a monitoring program must balance costs and objectives, and can best be made if the initial conditions in the watershed are known. A detailed baseline survey will identify areas which have desirable habitat and water quality, those which are degraded and the stream reaches where changes occur. The general nature of any stresses on the system can be described on the basis of the kinds of animals found at individual sites.
Sixteen sampling locations were selected within the western part of the Laurel Creek watershed to document baseline conditions and to determine the overall state or health of the basin (Figure 1). Sites were selected to represent all major sub-basins; det ermine the impact of impoundments; determine the effect of different stages of urban development, establish current conditions in the western portion of the basin; and assess the overall state of the basin. Five additional reference sites were selected on reaches of forested first and third order streams to determine the benthic invertebrate assemblages in "healthy" or undisturbed areas. The data collected from these reference sites were used to augment an existing reference database developed over the pa st 20 years. The results of this baseline survey will be used to design future monitoring programs for the Laurel Creek basin.
In the laboratory, each sample was emptied into a 100 Am aperture net, washed to remove most of the formalin, then examined under a dissecting microscope and at least the first 200 animals were removed to 70% ethanol for identification. Additional animals were removed until no obviously different taxa were found. Any remaining unsorted residue was scanned for large rare animals which were added to the sorted material. It took, on average, about 3 hours to sort the samples in this manner.
Invertebrates were identified and enumerated at the lowest practical taxonomic level using existing keys and literature. Identification required I to 4 hours per sample, depending on the variety of animals present.
The results were summarized for each station as:
The number of EPT should be high at undisturbed sites and decrease in proportion to increasing stress (e.g. siltation, nutrient enrichment or pesticide inputs). BI is calculated as the weighted average of the "tolerance" scores (based on distributions wit h respect to organic pollution in riffles of Wisconsin streams) of the species in a sample. Higher scores on BI reflect increasing organic loadings.
PMA expresses the degree of similarity between the composition of a sample and that of an expected community (the average composition of samples from physically similar reference sites collected at the same time of year). These expected communities have b een calculated from collections made throughout southern Ontario. In order to decide whether or not a site is significantly impacted, the lower confidence limit about the mean similarity of the samples contributing to the reference community is subtracted from the site PMA. Negative values indicate a significant difference from reference conditions. This also partially standardizes PMA values and facilitates comparisons among physically different streams. PMA is well suited for long-term monitoring becaus e it easily detects change toward, or away from, reference conditions.
The number of taxa sorted from individual samples ranged from 30 at Site 5 to 60 at Site 10 (Table 1). There was a strong correlation between the numbers of individuals identified and numbers of taxa. This suggests that if more indiv iduals were sorted and identified from any given sample, the number of taxa found there would increase. This is true to some extent, but we stopped sorting at about 200 animals only if no obviously new kinds were being found. The numbers of taxa per sampl e in Table 1 give a good indication of the relative diversity of the fauna among sites, but number of taxa is sensitive to sampling effort and provides no indication of the kinds of animals that are present.
EPT scores suggest that Clair Creek at site 16 is the most degraded stream reach that we sampled (Table 1). Sites 12 (Monastery Creek), 3 (Laurel Creek below Columbia Lake), 8 and 9 (Beaver Creek) also yielded small numbers of EPT, w hile large numbers were found at Sites 10, 6, 7, 14 and 1. It is worth emphasizing that no stoneflies were found at any site in the Laurel Creek basin.
The absence of stoneflies suggests at least some environmental degradation has occurred at all sites. This agrees with the assessment based on HilsenhoffÕs Biotic Index: the lowest score at any site, 5.03, would be evaluated as "good water quality with so me organic pollution" (Hilsenhoff 1987). The score of 8.58 at site 16 indicates very poor water quality. It should be emphasized, however, that the values of HBI reported here are not strictly comparable with those listed by Hilsenhoff (1987) because of d ifferences in sampling methods and the inclusion of molluscs and worms in calculating the index for the Laurel Creek survey. (When HBI was calculated using only arthropods (toe original method for this index), the worst sites appeared to be slightly less polluted and water quality ranged from "good"' to "fairly poor".) HBI scores were highest at sites 16, 12 and
The level to which animals are identified influences PMA scores, as shown in Table 1. In general, the sensitivity of PMA decreases as taxonomic categories are broadened. That is, the accuracy of site assessments increases with increa sing effort in identifying the organisms. However, accurate identification of many aquatic organisms to species is not possible with the taxonomic information available today, so identification to genus may be satisfactory.
In most cases the relative impact at a site was similar whether assessed at the level of species or genus, but overall water quality appeared to be somewhat better at the genus level. Exceptions were Beaver Creek site 9, Monastery Creek site 12 and Clair Creek site 13. Sites 5,7, 14 and 15 appeared to be in substantially better condition when assessed at the level of genus rather than species. Samples from the mainstream of Laurel Creek above Columbia Lake were not significantly different from reference c ommunities, nor were those from sites 7, 8 (Beaver Creek), 14 (Monastery Creek) and 15 (Clair Creek). The negative PMA score for Beaver Creek site 9 is somewhat surprising and deserves further investigation. The apparent improvements at the most downstream sites on Clair and Laurel Creeks are very encouraging.
|Site||# Animals||# of Taxa||EPT||Biotic Index||PMA - Species||PMA - Genus||PMA - Order|
Of the major tributaries, Beaver Creek appears to be in the best condition. Site 9 deserves closer scrutiny. The physical appearance of the stream at this point gives no evidence of recent disturbance but the sediment in the streambed is very loose and the benthic fauna is quite different from the expected community. Casual observations during several visits to Beaver Creek suggest that there are significant inputs of sediment downstream of site 7.
Monastery Creek in the vicinity of site 12 appears to be significantly degraded. The fauna is characteristic of areas where substantial amounts of fine sediments have been trapped. Most of the animals are also very tolerant of low oxygen, so there is also a possibility of organic pollution. Future monitoring should reveal whether this is a temporary result of recent construction in the immediate area, or an ongoing problem.
Evidence of the potential for recovery from disturbance during residential development can be seen in Clair Creek. Most of Clair Creek appears to be significantly degraded, especially adjacent to the newly developed area above Fischer-Hallman Road (site 16), but a much more balanced benthic community was found at Westmount Road.
Barton, D.R. 1996. The use of Percent Model Affinity to assess the effects of agriculture on benthic invertebrate communities in headwater streams of southern Ontario, Canada. Freshwater Biology 36: 397-410..
Hilsenhoff, W.L. 1987. An improved biotic index of organic stream pollution. Great Lakes Entomologist 20: 31-39.
Hynes, H.B.N. 1960. The Biology of Polluted Waters. Liverpool University Press, Liverpool, UK.
Lenat, D. R. 1988. Water quality assessment of streams using a qualitative collection method for benthic macroinvertebrates. Journal of the North American Benthological Society 7: 222-233.
Lenat, D.R., L.A. Smock and D.L. Penrose. 1980. Use of benthic macroinvertebrates as indicators of environmental quality. In: D.L. Worf (ed.), Biological Monitoring for Environmental Effects. Lexington Books, Lexington, MA: p.97- 112.
Novak M.A.. and Bode R.W. 1992. Percent model affinity: a new measure of macroinvertebrate community composition. Journal of the North American Benthological Society 11: 80-85.
Olive, J.H., J.L. Jackson, J. Bass, L. Holland, and T. Savisky. 1988. Benthic macroinvertebrates as indexes of water quality in the Upper Cuyahoga River. Ohio
Journal of Science 88:91-9X. Verdonshot P.F.M. 1992. Macrofaunal community types of ditches in the province of Overijssel (The Netherlands). Archiv fur Hydrobiologie, Supplement 90: 133-158.