By Ted Siler
Canoe Densities
On the reference date, 25 June 96, 40 canoes passed the sampling point at Wa Wa Sum with a high of 10 canoes at the 1600 hr and an average of 3.0 canoes hr1. On 29 June, 141 canoes passed the Wa Wa Sum sampling point with a high of 34 canoes at 1500 hrs and an average of 11.7 canoes hr1. This was an increase of 377% between reference and experimental sampling dates. Only the canoes passing the sampling point from 0900 to 2100 hrs were taken into account for these calculations, as this was the time when the majority of canoes were on the river. Any canoes appearing after 2100 hrs were always racing canoes and weren't considered because behavioral drift density increased after dark and would have obscured any canoeing effects.
On the reference date at Stephan Bridge (23 July 96) 55 canoes passed the sampling point with a high of 14 during the 1400 hr and an average of 4.0 canoes hr1 (0900-2100 hrs). During the experimental drift sampling on Saturday 20 July, 435 canoes passed the sampling point with a high of 98 canoes during the 1600 hr and an average canoe hr1 of 36.0. This resulted in 884% more canoes hr1 passing the Stephan Bridge site during the experimental sampling. Therefore, a gradaion of canoeing pressure was examined, with Stephan Bridge highest in canoe traffic.
Macroinvertebrate Drift Densities and Composition at Wa Wa Sum
Au Sable River macroinvertebrate drift densities varied throughout the sampling period when all sites were considered. Mean drift densities were at their second highest point during the April sampling at Wa Wa Sum, dropped to low values during June, and then increased during July to reach their highest densities during the Stephan Bridge samplings (Table 1). Drift densities for Wa Wa Sum on 27 April 96 followed the normal diel periodicity of drift noticed by other researchers (Waters 1962, Holt and Waters 1967, Pearson and Franklin 1968, Cowell and Carew 1976, Walton 1980, Benke et al. 1986, Moog and Heinisch 1991, and others). Drift densities were lowest (23.63 100m3) at 0700 hrs, just after sunrise, and highest (124.66 100m3) at the 0400 hr, two and a half hours before sunrise. This resulted in a decrease in drift rate after sunset of 527%. The mean nighttime drift density (86.5 100m3) was only 1.96 x greater than the mean daytime drift density (44.1 100m3). Others have also found differences in nightday drift levels. Benke et al. (1986) noted that night drift densities were 510 x higher than daytime densities in a 6th order river in Georgia. And a smaller difference of 2 to 2.5x the total daytime drift was found for nighttime drift densities in a 4th order mountain stream (Moog and Heinisch 1991).
Macroinvertebrate drift densities on the 25 June 96 reference date showed the normal diel periodicity of drift similar to the 27 April sampling. The mean nighttime to daytime drift density ratio was again small, with mean nighttime drift densities being only 1.8 x that of mean daytime densities. The highest mean nighttime value of 40.37 organisms 100m3 at 0400 hrs was 11.7 x the low daytime value of 3.45 organisms 100m3 at 0600 hrs. On 25 June the highest drift density peak was right before sunrise (0400 hrs) while the 29 June experimental sampling showed the highest pak in drift density after sunset (2300 hrs) with another high peak before sunrise (0500 hrs). The experimental date showed a higher night-day mean drift density ratio than the reference sampling. The mean daytime drift density value of 10.50 organisms 100m3 was 2.9 x lower than the mean nighttime value of 30.39 100m3. Both reference dates (27 April and 25 June) had the highest peak in drift density at the same time (0400 hrs) and lowest peak soon after sunrise. The experimental sampling exhibited the highest drift density after sunset with a smaller peak before sunrise. The April data also showed higher overall drift densities than the June sampling dates (Table 1), due to higher densities of Ephemeroptera and the presence of individuals in the order Plecoptera.
Mean daily drift densities for Wa Wa Sum during the June sampling dates were much lower than those previously found on the Au Sable River and in other similar sized rivers. Moog and Heinisch (1991) found drift densities of 140 100m3 in a fourth order mountain stream in Austria, while Eyman (1969), in the Au Sable River at Wakely Bridge, found a rate of 1,215,600 organisms day1 passing the sampling station. This rate was a 20 fold increase of the mean June rate (organisms day1) found at Wa Wa Sum. Drift density values from Wa Wa Sum in June were more similar to densities found by Cowell and Carew (1976) and Radford and Hartland-Rowe (1971) in smaller streams in Florida and Alberta, Canada respectively (Table 2).
The 25 June reference date contained 20 taxa in eight orders while the 29 June experimental nets captured 18 taxa in seven orders. On both dates the Ephemeroptera (39.6%) were the most abundant order captured, with the Diptera (28.5%) and Trichoptera (25.2%) following close behind. Together these three orders constituted approximately 94% of the drift on 25 June (reference) and 93% of the drift collected on 29 June (experimental). The Trichoptera, Coleptera, Simuliidae, and Chironomidae constituted approximately 75% of the drift numerically in a sith order Georgia river (Benke et al. 1986), while Berner (1951) found Chironomidae as the most numerous macroinvertebrate collected in the lower Missouri River. Waters (1972) stated that the taxa that have been recognized as most important quantitatively in the drift are the Ephemeroptera, family Simuliidae of the Diptera, Trichoptera, and Plecoptera in that order. The Ephemeroptera were also found to be most important in a small Florida stream, with the Coleoptera, Trichoptera, and Diptera second, third, and fourth respectively (Cowell and Carew 1976). The wide variety of organisms dominating the drift result from the variety of organisms found in the benthos of different rivers. If organisms prone to drifting are the most dominant in the benthos, they will also likely be the most prevalent organisms found in the drift. The principal taxa found in the drift of different rivers around the world varies due to the ecology of the river, it's location, and the seasonal factor associated with sampling.
The most common organism found drifting on both dates at Wa Wa Sum was the mayfly Baetis, which made up 22.3% of the mean total drift captured on the reference and 24.7% of the drift captured during the experimental sampling. Individuals in the genus Baetis are often reported as having high drift rates and are often the most common organism found in the drift during a 24 hr sampling (Waters 1972, Brittain and Eikeland 1988). The second most common taxa found in the drift during both dates was the blackfly family Simuliidae, it made up 17.5 and 16.7% of the mean total captured drift on the 25 and 29 June respectively. The Trichopteran Hydropsyche (14.6%) was the third most common drifter collected during the reference sampling, while on the experimental date the Chironomidae family (13.4%) was the third most abundant organism in the drift. Hydropsyche showed large decreases in nighttime % occurrence for the Wa Wa Sum experimental date when compared to the reference date, while Tricorythodes, Chironomidae, and Simuliidaehad a higher nighttime and total % occurrence on the experimental date.
Those organisms captured during the reference sampling but not the experimental sampling were the mayfly Isonychia, the Dipterans Empididae Hemerodromia and Tipulidae Antocha, the Odonate Calopterygidae Calopteryx, and the Hemipteran Corixidae Corixa. The Isonychia, Antocha, and Calopteryx were only found in one hour of sampling throughout the 24 hour sampling period. Except for Isonychia, which was represented by two individuals during 0300 hrs, Antocha (0900 hrs) and Calopteryx (1600 hrs) were only represented by one individual during the hour they were captured. The Calopteryx, a genus typically found in the detritus along the margins of lotic habitats (Merritt and Cummins 1984) was captured during the hour of peak canoe numbers (10) for the reference date. Although this number is small compared to canoe numbers on the experimental date, it was possible that this individual was forced into the thalweg by canoeing activity. The other taxa were captured during hours with no canoes present.
The 29 June experimental drift sampling captured three taxa that were not present during the reference sampling. The mayfly adult Hexagenia limbata Serville, the Coleopteran Hydrophilidae Hydrochara, and a terrestrial Homopteran family, Cicadellidae. Only three total Hexagenia limbata adults were captured during two sampling hours (0000 and 0100 hrs) and were most likely the result of dying adults landing in the water after ovipositing. A large number of mating adults were present during this sampling date (personal observation). Hydrochara was again represented by only one individual captured during the 2200 hr sampling. It is not known why the terrestrial Homopteran family on the 29 June sampling was found in the drift. These individuals were most likely not knocked into the water by canoers because one individual each was found during 2300, 0000, and 0600 hrs sampling and two individuals during the 0500 hr sampling. These were hours when no canoes wer on the river. Whitetail deer and small mammals were often moving in the riparian vegetation and the river itself during the nighttime hours (personal observation). It is possible that these terrestrial organisms were knocked into the river from this activity.
Macroinvertebrate drift densities for the 25 June reference were not significantly different (p< 0.15) than those for the 29 June experimental sampling date. Daytime values showed the largest differences between dates, but when comparisons were made, neither daytime (0900 to 2100 hrs) (p< 0.11) or nighttime (2200 - 0500 hrs) (p< 0.082) drift densities were considered significantly different. It was suspected that these dates were actually different, with drift densities on the experimental date lower, but were considered statistically the same because of high variability in the system and a low degree of freedom (2), making the test very conservative. It is hypothesized that with a higher number of replicates differences would have been seen. Although no differences were found, the total mean number of Ephemerella naiads, Hydropsyche larvae, and Brachycentrus larvae captured during the experimental sampling were less than half the number caught during the reference date. Baetis and Chironomidae were slightly lower on the experimental date, while the mayfly Tricorythodes increased five fold. This genus was the only taxon to show a marked increase during the experimental date. Merritt and Cummins (1984) describe Tricorythodes as a sprawler/clinger inhabiting depositional areas in lotic systems. Wave action caused by recreational canoeing, entered depositional areas of the river and agitated detritus and FPOM, forcing some of it to resuspend and travel downstream in the water column (personal observation). It is possible that any Tricorythodes naiads inhabiting these areas were resuspended with the detritus and carried downstream, resulting in slightly increased drift densities during the day. If these naiads were disturbed during the day, and drifted, but laded in unsuitable areas, it is likely that they drifted again in larger numbers at night in order to return to suitable habitats. Results of this scenario would be an explanation for the nine fold increase in Tricorythodes drift densities on the night of the eperimental date.
Macroinvertebrate Drift Densities at Stephan Bridge
The Stephan Bridge sampling dates exhibited much higher drift densities (five fold increase) than the April and June Wa Wa Sum samplings (Table 1), but were still below densities found in similar sized rivers (Table 2). The 23 June reference sampling exhibited a low of 35.26 organisms 100m3 at 1000 hrs and high of 289.0 100m3 at 2000 hrs. The 20 June sampling did not exhibit the large increase in drift densities before sunset displayed on the reference date. The mean daytime drift density of 81.56 100m3 and the mean nighttime density of 144.15 100m3 resulted in a night-day drift ratio of 1.8. Because the sun did not officially set until after 2100 hrs on the 23 July sampling, the 24 hr high for this date (2000 hrs) was included in the calculation of the mean daytime drift density, resulting in a low night-day ratio. When the 2000 hrs drift density is not included in the calculation of the daytime mean and considered a nighttime value, the result is a higher night-day drift density ratio of 2.4. When the altered night-day mean drift density ratio is used for 23 June an equal ratio was observed.
During the 20 July experimental sampling at Stephan Bridge, a low hourly drift density of 30.74 organisms 100m3 was found at 1100 hrs and a high density of 185.87 100m3 was exhibited at 0400 hrs. The mean daily drift density was 54.85 organisms 100m3 and the mean nightly drift density was 130.74 100m3, resulting in a nightday ratio of 2.4. Either due to date or site location, the Stephan Bridge sampling dates had a slightly higher mean night-day drift density ratio than the mean ratio for Wa Wa Sum dates.
The Stephan Bridge site exhibited a greater diversity of collected organisms than the Wa Wa Sum sit. Macroinvertebrates collected during the 20 July experimental sampling consisted of seven orders with 28 taxa, while the 23 July reference sampling contained eight orders represented by 25 taxa. Both dates contained members of the Ephemeroptera, Plecoptera, Trichoptera, Coleoptera, Diptera, Isopoda, and Amphipoda with two representatives of the order Megaloptera also found during the 23 July sampling. The 23 July displayed similar results with approximately 95% of the sample being represented by individuals in these three orders. On the reference date the Diptera were most common, constituting 58.7% of the sample. The 20 July found the Ephemeroptera as the most common order, representing 42.5% of the sample, with the Diptera close behind at 40.7%. Together the Ephemeroptera, Diptera, and Trichoptera orders constituted 94% of the collection. Similar to the Wa Wa Sum collections, the mayfly Baetis was the most common taxon collected (31.1% of the sample) during the 20 July sampling with the Chironomidae (24.4%) and Simuliidae (16.0%) second and third. On the 23 July the situation was reversed, with the Dipteran families Chironomidae and Simuliidae representing 29.4 and 28.9% of the sample respectively. The genus Baetis was a close third, constituting 25.5% of the sample. Baetis, Tricorythodes, and Brachycentrus exhibited large increases in % occurrence during the July experimental date while the Chironomidae and Simuliidae families were much lower.
The 10 taxa not found during the 23 July sampling but collected on the 20 July were the mayflies Isonychia, and Heptageniidae Stenonema, the Colepopteran Halipliidae Haliplus, the Trichopterans Limnephilidae Pycnopsyche, Philopotamidae Chimarra, Leptoceridae Oecetis, and Lepidostomatidae Lepidostoma, and the Dipterans Empididae Chelifera and Tipulidae Antocha. Six taxa were collected on the reference date but not the experimental and were represented by the Plecopteran Perlidae Paragnetina, the Coleopteran Dytiscidae Dytisca, the Dipteran Tipulidae Tipula, the Megaloptran Corydalidae Nigronia and the Trichopterans Rhyacophilidae Rhyacophila and Helicopsychidae Helicopsyche. When both dates were combined the order Trichoptera was the most diverse, with 11 families represented, and the Diptera were the most numerous, constituting almost half of all organisms collected at Stephan Bridge.
As was the case at Wa Wa Sum, no large numbers of any taxa were noticed in one sample date and not the other. Most taxa differences consisted of only a few individuals collected during the 24 hr sampling series. Most likely these individuals are part of the constant drift (Waters 1965) resulting from accidental dislodgment from the substrate and not normally found in considerable numbers in the drift. The taxa differences noted on these dates constituted only 0.29% of the total drift collected at Stephan Bridge and 1.20% of the total drift at Wa Wa Sum. It was hypothesized that the appearance and timing of individuals on the reference date and not the experimental date, and vice-versa, were purely coincidental and not the result of any canoeing effects.
Twenty-four hour macroinvertebrate drift densities for 20 and 23 July 96 at Stephan Bridge were found to be significantly different (p< 0.021), with the 23 July reference sampling lower than the 20 July experimental sampling. When daytime (06002000 hrs) drift densities were compared between dates it was found that values were considered significantly different (p< 0.05) with the experimental date lower, while nighttime (21000500 hrs) drift densities were not (p< 0.16). Tricorythodes naiads and Brachycentrus larvae increased by two fold during the experimental sampling, while Hydropsyche and the Dipteran families Chironomidae and Simulidae all showed fairly large decreases. The cased Brachycentrus attaches its case to rocks, aquatic macrophytes, and large woody debris in such a way as to use its extended hairy metathoracic legs to collect suspended organics in the water column (Merritt and Cummins 1984). This food collecting position subjects the lrvae to greater increases in catastrophic drift from substrate disturbance. Although the cases were firmly attached to the substrate (personal observation), it is possible that canoe action (paddles, canoes, or the operators themselves) disturbed the benthos or aquatic vegetation and removed the larvae and caused them to drift.
These data suggest high numbers of canoes on the river may actually be suppressing drift of most macroinvertebrate taxa during the day instead of increasing total drift catastrophically, as was originally hypothesized. Although no significant differences were found for the Wa Wa Sum sampling dates, a trend of lower drift densities during the experimental date, coupled with the lower drift densities found at Stephan Bridge, support this hypothesis.
Large daily differences in drift were found in a small coastal British Columbia stream, but Williams (1980) attributed most of the differences to a few drifting taxa that did not normally constitute a large portion of the drift. Studies examining catastrophic drift have found increases in macroinvertebrate drift densities, taxa richness, and biomass due to benthic disturbance (Waters 1962, Eyman 1969, VanHouten 1986). But it has also been noted that during increased disturbance some organisms may actually burrow into the substrates to escape abrasion by fine sediments (Culp et al. 1986) or decrease the chances of entering the water column and being swept downstream (Elliot and Bagenal 1972).
Macroinvertebrate Regression Analysis
When regression analyses were performed for the Wa Wa Sum (rsq = 0.066 p< 0.186) and Stephan Bridge ( rsq = 0.017 p< 0.565) sampling dates, it was found that no linear relationship existed between macroinvertebrate drift densities and log canoe abundance + 1. The original hypothesis stated that a linear relationship would exist between drift density and canoe abundance. Because this relationship was not present it was hypothesized that if canoes are affecting macroinvertebrate drift densities it is most likely a loclized event, with short drift distances, instead of a river wide occurrence.
It was noted by McClay (1970) that when organisms were dislodged by substrate disturbance, mean drift distances were only 10.7 m, with a range of 0.5 to 19.3 m, and a maximum distance of 45.7 m. Elliot and Bagenal (1972) found similar distances in a stony stream in the English Lake District and estimated catastrophic drift distances were from 2 33 m. Waters (1965) hypothesized from blocking experiments that nightly drift distances were 50 60 m, and that organisms moved downstream in a saltatory fashion. On the other end of the spectrum, some species of aquatic insects have been known to drift > 500 m when physical conditions of the stream bottom did not favor reattachment (Vinikour 1981). This is most likely not the case in the Au Sable, as bottom substrates consisted of good macroinvertebrate habitat (gravels and cobbles) throughout the study sections.
From these short drift distances it was suspected that any organisms that were forced into the water column by canoes in the Au Sable River did not drift far before returning to the substrates. With large drift distances not present, canoers would have had to dislodge organisms from within short distances (< 50 m) of the drift nets, and do so at a time when the nets were in position to detect any peaks in drift density as a result of canoeing activity. Evidence for this situation in the Au Sable River was provided by Eyman (1969). A canoe capsized approximately 10 m upstream from his drift nets, resulting in a considerable disturbance to the stream bed and a large peak in macroinvertebrate drift density.
VanHouten (1986) suggested that large numbers of canoes or fisherman causing catastrophic drift may lower macroinvertebrate densities in a river or stream enough to affect fish growth and abundance. The opinion of this author is that the number of animals that are forced into the drift during these activities, although they may exhibit some increased mortality (Hynes 1970, Wilzbach t al. 1988), is so minimal when compared to benthic population numbers that to suggest a decrease in macroinvertebrate populations to a point where insectivorous fish are affected, is likely false.
It has been calculated for some streams that approximately < 0.01% of the benthic macroinvertebrate population may be found in the drift at any one time (Elliot 1971, Williams 1980) and Elliot and Bagenal (1972) found only a small loss from the total benthos (5%) when they induced catastrophic drift in a stony English Lake District stream. This percentage of animals drifting compared to the benthic population is small enough to imply that a small increase in drift numbers due to catastrophic drift would not cause a great reduction in macroinvertebrate densities. Because areas of stream bottom that have been denuded by disturbance are known to recover from individuals drifting in from upstream (Waters 1965, Gore 1979), any small area disturbed by fishermen or canoers is most likely colonized rather rapidly, thus not exhibiting a reduced benthic population. Hynes (1970) suggested that if excess production does not accumulate on the stream bottom but instead enters the drift and becomes food for fish or is otherwise lost (Waters 1966), the relationship between benthic biomass and fish production is a loose one at best.
In conclusion, the large amounts of recreational canoeing that is present during the summer months on the Au Sable River is not causing significant amounts of macroinvertebrate catastrophic drift. However, it does seem to change the composition of the drift, increasing the occurrence of some taxa, while decreasing the occurrence of others. What affect this might have on overall macroinvertebrate densities in the Au Sable is unknown. Although macroinvertebrate drift densities recorded during the summer of 1996 are lower than drift densities reported in the literature from rivers of similar size as well as earlier years on the Au Sable, this should not automatically result in a conclusion that drift densiies are low enough to be detrimental to the Au Sable fish populations. When Eyman (1969) conducted drift studies on the Au Sable in the late 1960's, the river was still being fertilized by output from the Grayling sewage treatment plant. Artificially high nutrient concentrations almost definitely resulted in the high drift densities recorded during his study. A better test would be to obtain present day macroinvertebrate drift density values from rivers in the same locality as the Au Sable, such as the Manistee, Pine, or Pere Marquette rivers, and use them as a comparison to those found here.
It is my recommendation that yearly or biyearly drift sampling be conducted on the same sites and dates used in the present study. By comparing macroinvertebrate drift densities between years, as well as with MDNR fish shocking surveys, any link between drift densities and fish production can be established or disregarded as a possible reason for the decline in fishing success in the upper main stem of the Au Sable River.
Results of the additional three sections of this research (suspended solid concentrations, detritus movements, and algal drift) will be presented and discussed in a later issue of this newsletter. RWOL
© Copyright 1998 - , Anglers of the Au Sable, Inc. All rights reserved. Last modified: February 11, 2002
