A Report to the
Ten Towns Great Swamp Management Committee

December, 2004

A return to normal rainfall levels over the past two years is matched by a clearimprovement in macroinvertebrate community quality.  Among the 17 sampling sites, 3498 individuals representing a total of 136 species were observed.  In all but 3 cases, 2004 B-IBI values are equal to or exceed 2003 values.  In fact, B-IBI values in 14 of the 17 sites equal or exceed maximal scores observed over the last 4 years.  Positive changes in community characteristics at more stressed streams, Black, Loantaka, and Great Brooks, included declines in the degree of dominance by a few species, some increases in the total number of taxa present, and, somewhat peculiarly, substantial increases in the proportion of the fauna comprised of predators.

Community quality metrics, B-IBI and Becks Index, correlate with most variables, but especially with those that characterize habitat conditions.  Strong negative correlation is also seen between B-IBI measures and variables such as total dissolved substances (TDS) and temperature, while the correlation with dissolved oxygen and stream gradient are positive.  For the second year in a row, highest TDS values occurred not at the headwaters of Loantaka Brook but farther downstream.  Another TDS oddity is the dramatic decline in TDS readings at the uppermost sampling site on Great Brook, GB5, just below Foote's Pond.

Some 82% of the variability in B-IBI measures averaged over 2000-2005 can be explained by variations in mean habitat value for the same period.  Mean dissolved oxygen accounts for 54.3% of B-IBI variability, while mean total dissolved substances account for 52.3%.  The negative correlations between mean temperature and mean turbidity and B-IBI score are less predictive.  No correlation exists between pH levels and community quality.  From the distribution of three major watershed habitat modifiers, i.e., sewage treatment plants, impoundments, and parking lots/detention ponds, it is clear that all sites in poorer quality streams experience at least one of these sources of stress.

Refer to Stream Summaries at the end of this document for more details.

The now familiar pattern of broad ranges in TDS and Habitat Value underscore the continuing cause for concern regarding the impact of human disturbance across the Great Swamp watershed along a west (lower) to east (higher) gradient.  As mentioned in previous reports (e.g., Pollock, 2003), lower quality (i.e., eastern) sites have shown greater variability over the years (see Table 04-2) than have higher quality (i.e., western) sites.  It makes sense that these measures, emphasizing erosion and sedimentation, vary more in the impacted streams.  This year improved conditions relative to sediment build up were seen at BB1, LB2, and PR1.

Significant changes in overall conditions at LB2 and GB4 were obvious upon on-site inspection.  At LB2, striking erosive enlargement of the pool just downstream from the sampling site had occurred during the preceding year.  Perhaps the same water flow event(s) that caused this change was responsible for flushing sediment accumulation out of the LB2 site, mentioned above as improved.  We may be seeing the downstream consequence of this (and probably other) sediment erosion in building sediments at LB1 downstream.  Human intervention at GB4 has completely altered habitat conditions.  The stream bed has been widened at this location, vegetational cover has been decreased, but cobble-sized substrate has increased, leading to an overall in-stream habitat quality upgrade.

In Figure 04-2, B-IBI values from 2004 samples are compared to samples from 2000-2003.  For reference, the mean scores for each site are shown as are scores from 2003 and indications of the high and low range of scores.  B-IBI values in 2004 are equal to or exceed 2003 values in all but 3 cases (each of which show only slight (2 point) declines).  In fact, B-IBI values in 13 of the 17 sites equaled or exceeded maximal scores observed over the last 5 years, see Figure 04-2.  Eight sites have scores that differ from 2003 by 4 points (BB1, LB4, GB3, GB4, PR1), 6 points (LB1 & PR2), or by as much as 10 points (LB3).  Reasons for these and other changes can be identified from among the eight components used to calculate the B-IBI value in Table 04-4, although it may be easier to visualize these specific differences by referring to summary Table 04-5.  In Table 04-4, 2004 data points that are significantly different than 2003 data (enough to change B-IBI scores) are highlighted in underlined, bold, italics.

Positive changes in community characteristics at more stressed streams, Black, Loantaka, and Great Brooks, included declines in the degree of dominance by one or two species, some increases in the total number of taxa present, and, somewhat peculiarly, substantial increases in the proportion of the fauna comprised of predators.  As might be expected, the stressed streams showed little improvement in the numbers of sensitive organisms present, which limits even their improved scores to the "poor" category (see Fig. 04-2).  The reverse is true of the less-stressed streams where changes resulted from a decline in stress-tolerant species that were presumably displaced by increases in sensitive mayflies and stoneflies.

Even though labeling categories of "very poor", "poor", "fair", and "good" have been assigned in rather arbitrary fashion to the numeric B-IBI scale, they have been consistently applied to this study over the years.  In 2004, four sample sites remain "good" (PB2, PB3, PR3, IG1), the two lower Passaic River sites improved from "poor" to "fair" and join PB1 there.  LB2 and GB2 remained "poor" but all the remaining Loantaka Brook sites improved from "very poor" in 2003 to "poor" in 2004.  Likewise, Great Brook sites GB3 and GB4 also rose to "poor" in 2004.  This left only BB2 and GB5 (which dropped by only 2 B-IBI points) in the "very poor" category.

In June 2004, we appear to be seeing Great Swamp stream macroinvertebrate communities at their best.  Improvements both last year and this year appear to reflect the recovery of many of these sites from lower and slower water flows with warmer temperatures and less available dissolved oxygen associated with a multi-year period of lower than normal rainfall.

In Table 04-6, I have used a box and bold type to highlight statistically significant correlations (coefficients > +/- 0.482 at df = 15, p<0.05) between the two biotic indexes calculated here, Beck and B-IBI, and environmental variables monitored on June 11, 2004.  First, it is reassuring that the metrics, B-IBI and Beck Index, are highly correlated with one another ( r = 0.952).  They also both show matching correlations with most other variables.  The measures show high correlation with changes in most of the various habitat assessments, including their Totals and subtotal Habitat Values.  Strong negative correlation is also seen between B-IBI measures and variables such as total dissolved substances (TDS) and temperature, while the correlation with dissolved oxygen and stream gradient are positive.  Several variables show strong correlation with one another.  Stream bank stability is negatively correlated with TDS levels, and positively correlated with stream gradient (overall steepness) and with sedimentation and the embeddeness of substrate elements associated with it.

Over the years, one-shot measurements of environmental variables taken on a single date during the sampling period, e.g., Table 04-1, have proved useful.  However, it is also instructive to smooth out individual annual variability by repeating the analysis using mean values for B-IBI and for variables averaged from five years of available data, 2000-2004, with results shown as a table of correlation coefficients, Table 04-7.  Comparisons between correlation coefficients from averaged data (Table 04-7) and from June, 2004 data alone (Table 04-1) show even stronger correlations in the former.  Community quality, expressed as mean B-IBI, shows significant correlation with nearly all the environmental variables surveyed, with the exception of the degree of riparian cover, pH, and a series of descriptors of the size of the stream at the sampling site, i.e., order, width mean depth, mean velocity, and discharge volume.

We can explore further the relationship between community quality (B-IBI) and variables in Figures 04-3 through 04-7.  In each case a best-fit regression line has been added along with its formula and a calculation of the Coefficient of Variation, R² (estimates the proportion of changes in the dependent variable, i.e., B-IBI or community quality, that can be explained by matching changes in the independent variable, i.e., the environmental measure in question).  Fig. 04-3 shows a very strong positive correlation between mean B-IBI score and matching mean habitat value scores.  Some 82% of the variability in B-IBI can be explained by variations in habitat value.  Mean dissolved oxygen accounts for 54.3% of B-IBI variability, Fig. 04-4, while mean total dissolved substances account for 52.3%, Fig. 04-5.  The negative correlations between mean temperature and mean turbidity and B-IBI score are less predictive, Fig. 04-6 and 04-7 respectively.  No correlation exists between pH levels and community quality, Fig. 04-8.

Vertical lines have been added to several figures to indicate values above or below which the particular variable appears to play little or no role in determining community quality.  For example in Fig. 04-8, pH levels between 7.3 and 7.7 are associated with a wide range of B-IBI scores, by levels above or below that range are associated with lower quality communities only.  Similarly, conditions including turbidity less than 2 FTU, temperature less than about 20 C, and total dissolved substances less than about 170 ppm appear to have little influence over community quality, while conditions higher than these limits are associated with poorer quality communities. Community quality is more evenly arrayed along gradients of habitat value, Fig. 04-3, and dissolved oxygen, Fig. 04-4.

Three major habitat modifiers that appear to most adversely influence immediately downstream sites within the Great Swamp watershed, are sewage treatment plants (impact BB2 and LB3), impoundments (impact BB1, LB2, GB3, GB5, PB3, and PR1), and parking lot runoff/detention ponds (impact LB4, GB4).  These influences are shown along the x-axis in Figure 04-2.  There is a clear distinction between B-IBI assessed macroinvertebrate community quality at struggling Black, Loantaka, and Great Brooks and much higher quality at Primrose Brook and Passaic River sites.  From the distribution of the three habitat modifiers, indicated by symbols surrounding site labels, it is clear that nearly all sites in poorer quality streams experience one of these sources of stress.

1. Monitoring of these 17 sampling sites should be continued.  The longer the run of data regarding these key locations throughout the watershed, the more clearly we will be able to discern the difference between "normal" variations and those indicating significant change.  I will submit a proposal for the extension of this monitoring program to the Ten Towns Great Swamp Watershed Management Committee.
    
2. Under the generally improved rainfall conditions across the watershed, most sites currently host perhaps the best quality macroinvertebrate communities that we can expect.  Only two sites remain in the "very poor" category.  Of the two, GB5 should return into the "poor" range as conditions in the Foote's Pond impoundment just above it are improved.  We should watch for this to occur and seek further explanations if it does not.  The environmental context for the sewage treatment plant influenced BB2 may never permit the community at this site to improve.
    
3. Visually conspicuous sediment buildup was observed at LB1, while evidence of strong stream bank erosion was seen at LB2.  This stream in particular appears to be a good candidate for stream bank stabilization efforts to curb the influx of sedimentary materials in the stream.

Barbour, M.T., J.Gerritsen, B.D.Snyder, and J.B.Stribling.  1999.  Rapid Bioassessment Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates and Fish, Second Edition.  (EPA 841-B-99-002).  U.S.  Environmental Protection Agency; Office of Water; Washington, D.C.

Karr, J.R.  and E.W.  Chu.  1999.  Restoring Life in Running Waters: Better Biological Monitoring.  Island Press, Washington, 206 pp.

Pollock, L.W.  2000.  The Macroinvertebrate Communities of the Great Swamp Watershed.  A Report to the Ten Towns Great Swamp Management Committee.  General Introduction and Methods: 2000 and Subsequent.  12 pp., Tables & Figures.

Pollock, L.W.  2002.  The Macroinvertebrate Communities of the Great Swamp Watershed, June, 2002: Results.  A Report to the Ten Towns Great Swamp Management Committee.  11 pp, 6 Tables, 6 Figures.

SCS Water Quality Indicators Guide: Surface Waters.

Figure 04-2.  B-IBI values for 17 sampling sites within the Great Swamp watershed for June, 2004 & June, 2003 compared with the mean and range of values from 2000-2004.

Figure 04-3.  Mean B-IBI vs. mean habitat values for data from 2000-2004.

Figure 04-4.  Mean B-IBI vs. mean dissolved oxygen for data from 2000-2004.

Figure 04-5.  Mean B-IBI vs. mean total dissolved substances for data from 2000-2004.

Figure 04-6.  Mean B-IBI vs. mean temperature for data from 2000-2004.

Figure 04-7.  Mean B-IBI vs. mean turbidity for data from 2000-2004.

Figure 04-8.  Mean B-IBI vs. mean pH for data from 2000-2004.

All Figures.     All Tables.     Appendix 1.