A ~1000 Year Record of Sedimentation in Collins Lake as Evidence of Local Storminess and Flooding on the Mohawk River (New York)

Ian Robert White
Donald T. Rodbell

Geology Department
Union College
Schenectady NY 12308 (USA)
Suggested Citation:

White, I.R., 1999, A ~1000 Year Record of Sedimentation in Collins Lake as Evidence of Local
Storminess and Flooding on the Mohawk River (New York); Senior Thesis, Union College Geology
Department. 79 p.



Collins Lake is an oxbow lake situated on the floodplain of the Mohawk River in Scotia, New York and receives river sediment on a regular basis, perhaps every 10 years. A 582 cm long core, Core A, from the deep basin of the lake reveals laminated stratigraphy containing two main types of layers, pink silt and brown silt. Grain size trends within the layers and provenance analyses using carbon content and color indicate that pink layers are suspended sediment from the Mohawk deposited by overflows and brown layers are mobilized sediment from the basin cliffs or bedload from the Mohawk deposited by underflows. Radiocarbon dating indicates a basal age of ~1000 yr.BP from Core A, yielding a sedimentation rate of ~0.57 cm/year. A 14C date of 710 +/- 50 yr.BP obtained from 248-252 cm in a similar core from Allison, 1995 correlated to Core A by matching magnetic susceptibility curves yields a sedimentation rate of 0.63 cm/year. Due to lack of a good age model, a constant sedimentation rate of 0.6 cm/year was adopted for Core A. Avulsion of the paleochannel to the Mohawk and the onset of the sediment record occurred at ~1050 AD with a period of intimate connection with the river for the next 100 years, as evidenced by frequent pink layers and an overall coarser background sediment. Co-variation of pink and brown layers indicates a period of increased storminess from ~1150 to ~1590 AD that may be the result of the Little Ice Age. The most recent 200 years of sedimentation are virtually devoid of lamination because of the construction of dikes and increased bioturbation. Carbon and grain size data, however, suggest a second wet period from 1840 to 1915 that corresponds with a period of low salinity recorded in Chesapeake Bay sediments. (Note that this web summary does not contain figures.)


Lacustrine sediment records may provide clues to the origin and evolution of a lake. In particular, trends in the sediment properties, type of sediment, and nature of sedimentation can uncover climatic and anthropogenic changes that have affected a lake. Some lakes, however, are more interesting than others. For instance, a lake located on the floodplain of a river is in a unique setting because presumably, if its drainage basin is small and it has no major inflow, the lakeÕs main source of sediment is the river. Thus, the rate of sedimentation in the lake would be abnormally high for the size of its watershed, and its sediments would record all of the floods on the river that were of a great enough magnitude to reach the lake.

If conditions are right, a sediment core from this type of lake can be used to date and characterize flood events on the associated river and changes in magnitude and frequency of storm events through geologic time. If it can be proven that certain layers correlate to flood events and there is datable material in the core to provide age control, then essentially a chronology of river flooding since the birth of the lake may be obtained. In order to be preserved, floods have to be high enough to spill into the lake, sediment load has to be great enough to be recorded, and the flood layer has to be undisturbed (i.e. bioturbation), but otherwise the record is complete. Other records of flooding on rivers are short, like historical records, or incomplete, like floodplain sediments. Ultimately, changes in flood magnitude and frequency, may be linked to climatic and anthropogenic effects on the lake.

The focus of this thesis is to test the hypothesis of Allison (1995) that the stratigraphy of sediment in Collins Lake contains evidence of large magnitude floods on the Mohawk River. The goal, in essence, is to assemble a flood record from a sediment core from the deep basin of Collins Lake, New York, which is a floodplain lake of the nature described above.

Previous Applications of Lake Sediment Cores to Flood Chronologies

Caldwell and FitzGerald (1995) characterized the geomorphology, hydrology, and sedimentation of lake-outlet deltas in Maine. These deltas (Figure 1) are found in lakes at the mouths of outlet streams that lead to larger trunk streams. They found that outlet ?delta sedimentation occurs during flood events when the stage of the river exceeds that of the nearby lake and creates a flow reversal in the outlet stream (Caldwell et. al., 1995). Also, when looking at sediment cores of the delta they inferred that coarser sand represents sediment carried into the lake via an outlet stream during a flood reversal event, and overlying finer silt is deposited during slack water conditions occurring after the lake and the river equilibrate or as background sediment between flood events. (Caldwell et. al., 1995)

Rasanen et. al. (1991) investigated the viability of using 14C dating of floodplain lake sediments to create palaeoecological models. They concluded that the dates obtained must be viewed with caution and the climate oscillation must be considered tentative because denudation and lateral erosion of older alluvium recycles old organic material. Additionally, because the sediment of floodplain lakes is dominated by allochthonous material from the river, much recycled organic material gets into the lake. Thus, dates from lake sediments provide age maximums, but are quite possibly erroneously old. To help reduce error, dating should be based on small sediment samples with a high concentration of small organic fragments, rather than large samples with large clay and silt content and low organic content (Rasanen et. al., 1991).

Previous Studies on Collins Lake

Because of its interesting characteristics and proximity to the campus, Collins Lake has been the subject of a number senior theses at Union College, not only in geology, but in the other natural sciences and engineering. In 1994 the Union College Geology Department began its research on the lake by taking a 4.7 m sediment core from its deep basin, which provided the basis for three senior theses. Howk (1995) examined charcoal fragments in the core to establish a record of fire frequency in the area. Dlubac (1995) used pollen grains in the core to reconstruct past vegetation in the area, and linked trends in exchangeable metal abundances to human activities, and Allison (1995) analyzed the stratigraphy and sedimentalogy of the core. Allison asserted that coarser layers in the bottom half of the core are turbidites induced by high magnitude flooding of the Mohawk River. This was the first study directed at assembling a chronology of flooding on the Mohawk River that pre-dated written history.

Ruggiero (1999) used seismic profiling and geomorphologic characteristics of the lake to argue that it originated as an oxbow of the Mohawk River. Also in 1999, students John Gara, Karen Luey, and Jason Lederer, and Professor John Garver completed characterizations and chronologies of flooding and ice-jamming on the Mohawk River. Additionally, van der Bogert (1978) presented a simple chronology of Mohawk River Flooding. Finally, Carl George, a Union College professor of biology, has devoted much of his research to the many interesting facets of Collins Lake including anthropogenic history, archeological studies, and monitoring of the extensive restorational dredging in the mid 1970Õs.


To date, there have been no studies of the full sequence of Collins Lake sediments in the deep basin. The main objective of this study is to test the hypothesis of Allison (1995) that the stratigraphy of sediment in Collins Lake contains evidence of large magnitude floods on the Mohawk River. This objective will be accomplished in three ways: 1) create an age profile of the entire sediment sequence, including a basal age of the lake; 2) stratigraphically characterize the core from top to bottom; and 3) determine the nature and source of relatively coarse layers and layers of outstanding color. Ultimately, the hope is to gain an extended record of major flooding on the Mohawk River and relate this to regional climate change and anthropogenic activities in the region.


Collins Lake is located in the heart of suburban development around Scotia as the center piece of Collins Park. Its history has been affected by human activity virtually since its conception ~1000 years BP, when Native Americans would have inhabited its banks (George, 1999). Major anthropogenic influence on sedimentation in the Lake, however, first occurred with European settlement, bringing massive deforestation. By the early 1800Õs deforestation had caused decreased slope stabilities in New England and higher sedimentation rates that lasted at least until the 1880Õs when forests began to take hold again (Bierman et al., 1997).

On a more local scale, construction of the Schonowee Dike in 1804 (George, 1999) to the south of the lake as a carriage route from the Burr Bridge would have raised the flood stage that the Mohawk had to reach in order to overtop the dike. Previous to this construction, the river and the lake may have been more intimately connected (George, 1978). The dike was raised to its present elevation five meters above mean river level in 1835. One year after the initial construction on the Schonowee Dike in 1804, the construction of abutments for the Burr Bridge to the southeast of the lake (Figure 1) may have introduced a new ice-jam point during spring breakup (Gara, 1999). The bridge was later condemned, but today the abutments still remain (Gara, 1999).

The current level of the lake is 1.5 meters higher than it was in the early 1800Õs when its level was first recorded. This rise in lake level and subsequent increase in the lakeÕs aerial extent can be attributed to several factors. The first is the construction of the Schonowee Dike, which delineated and deepened the lake (George, 1993). In 1914, the construction of the dam at Vischers Ferry, as part of the new Mohawk River Barge Canal, caused the local water table to rise about one meter enlarging the lake and submerging the adjacent wetland (George, 1993). Finally, in 1946, the dam that Washington Avenue runs atop on the east-end of the lake was constructed with a weir and flapper valve to regulate the lake level to its current level of 214Õ a.s.l. (George, 1993). Around the same time, dredge spoils from the Mohawk associated with the Barge Canal were introduced to the lowland to the south of the lake between the lake and Schonowee Dike, further isolating the lake (George, 1993).

Finally, the lake has been severely impacted by the introduction of foreign plant species. The water chestnut was introduced into the lake prior to 1884 and peaked in the 1930Õs and 40Õs when a major campaign to control the weed began, including copper sulfate, hand- and machine-pulling, and cutting boats. The campaign was successful, but it was only to be followed by an invasion of the curly-leaved pondweed which contributed to eutrophication of the lake. To control productivity in the lake, surface sediment was dredged from the north and south sides of the island in 1977-78 amounting to the removal of 52,000 cubic meters of sediment from 10% of the lake bottom (George, 1982). The sediment was placed in a sedimentation basin in the wetland to the south of lake. A similar project was completed from 1989-94.

Regional Setting

Collins Lake is located on the floodplain of the northern bank of the Mohawk River in the Village of Scotia, in Schenectady County, New York (Figure 2). The lake has an area of 0.13 km2, a maximum depth of 9.75m, a mean depth of 3.8m, and a pronounced summer thermocline at 4 to 5m (George et. al., 1993). The west end of the lake is a deep basin, while the east end is a shallow flooded marshland, and nested between these is an oblong island trending E-W (Figure 3). The island is owned by the Nature Conservancy. The lake is ~500m from the river, separated to the south by a flat, man-made dike, the Schonowee Dike, which is ~5m above mean river level and ~4m above mean lake level. Situated on the flat is Collins Park, home to baseball fields, an ice rink, and a small public beach made of imported sand. Adjacent to the north/northwest bank of the lake is a steep bluff that outlines the lake, which is inferred to be a paleo-cutbank of the Mohawk River (Ruggiero, 1999). Flow in the lake is to the east, where it drains through a small culvert with a flapper valve into Collins Creek and finally into the Mohawk River (Figure 3), which also flows to the east. From personal observation it is know that Collins Creek backs up its channel and flows into the lake whenever the stage of the river is higher than that of the lake (Figure 2). A rise in river stage of ~1m is sufficient to cause flow into the lake, and this occurs perhaps annually. Additionally, when the stage of the Mohawk River is above the elevation of the Schonowee Dike, which requires a rise in river stage of ~5m and may occur once every two decades, flood waters have been observed coming directly over the dike from the Mohawk into the lake (Figure 2).

Bedrock and Surficial Geology

Bedrock directly below Collins Lake consists of Ordovician Flysch of the Schenectady Formation. The Schenectady Formation in Scotia is a ~1km thick sequence of medium- to thick-bedded graywacke turbidites alternating with deep marine shale. A thick section of Paleozoic sedimentary rocks, including carbonates, underlies the Schenectady Formation (Gara, 1999).

The steep bluff to the north of the lake consists of well-sorted sands of the Mohawk Delta. The delta formed on the west side of Glacial Lake Albany during the last deglaciation (Allison, 1995). The island in the center of the lake is composed of coarse river gravels. The dikes on the east and south sides of the lake are made up of imported fill, and the overburden to the southeast of the lake is composed of dredge spoils from the Mohawk River and Collins Lake.


In the winter of 1999, during a Lakes and Environmental Change coarse taught by Paul Gremillion and Donald Rodbell, two long cores were taken from Collins Lake with a square rod piston corer. The first, Core A (582cm), was taken from the deep hole to the west of the island, and the second, Core 2 (~500cm), was taken closer to shore, north of Core A (Figure 3). Both cores were split into two halves, a working half and an archive half, and all samples to follow were taken from the working half. Before the cores were unwrapped, both were run through a Bartington magnetic susceptibility meter to test their ferrimagnetic abundance. A data point was recorded for every two centimeters. Then, 1cm3 volumetric samples for bulk density and total carbon were taken approximately every 10cm in both cores.

Bulk samples were taken from organic rich layers or layers with macro organics for radiocarbon dating in Core A at 303cm, 386.5cm, 448cm, 530.5cm, and 582.4cm below the water/sediment interface, and in Core 2 at 269cm, 359cm, 422cm, and 459cm below the water/sediment interface. The samples were then sieved through a 250 um screen and all remaining organic material was sent to University of Colorado at Boulder for preparation and Woods Hole for AMS Radiocarbon Dating.

A complete stratigraphic column of Core A was created with emphasis placed on brown and pink silty layers, and contacts. Extensive volumetric sampling of Core A (~10 samples per drive) was completed. These samples, along with small bulk samples taken through individual layers, were analyzed with a Coulter Laser Diffraction Unit to characterize grain size variations. All samples were analyzed for total carbon with a UIC Autofurnace at 950? C and a UIC Coulometer. Certain samples were also analyzed for inorganic carbon using a UIC Acidification Unit and a UIC Coulometer.


Investigation of the stratigraphy of core A reveals three sediment types of importance to this study. The first is a fairly organic-rich brown silt and the second is organic-poor pink silt, both of which occur as distinct layers, ~0.5 to 6 cm thick. The brown silty layers exhibit basal scouring in some instances and both layer types show sharp contacts and gradational grain-size trends. The abundance of both layer types increases down profile with a drastic change from stratigraphy with virtually no silty or pink layers to stratigraphy with very frequent layers, which occurs at ~250 cm (Figure 4). Additionally, down-profile, brown silt layers and pink silt layers begin to appear as pairs where a thinner pink layer overlies a brown layer. The remainder of the sediment includes brown mud, gray mud, gyttja, or some combination of the above including pink mud and brown silt. This sediment is characteristically massive, undifferentiated, and heavily bioturbated, it will be referred to as background for the remainder of the paper.

Certain discrete grain size samples that neighbor each other are grouped together as layers, which are labeled A through L on Figure 6. Grain size data (Appendix 1) for background sediment generally show a grain size peak around 5 ?m and a characteristic shoulder from 12-20 ?m (Figure 5). Both pink layers and brown silty layers are coarser than the background, but to varying degrees (Figure 6). Silt layers frequently contain grains up into the fine sand range and show characteristic double peaks on the distribution curves (Figure 7). Additionally, silt layers generally show coarsening upward sequences where both discrete samples have the characteristic dual peaked curve (Figure 8). Pink layers, however, exhibit characteristic smooth bell curves barely reaching the coarse silt range and show fining-upward sequences (Figure 9). In conjugate pairs with an underlying silt layer, pink layers have a slightly modified grain size curve (Figure 10). Finally, the coarse component of both silt curves and background sediment curves are greatly exagerated for samples below ~480 cm (Figure 11). In profile, statistical grain size data (Appendix 2) show high variation in coarse-grained material which is far more abundant below ~250 cm down profile (Figure 12). Presence of fine-grained material is highly variable, but does not show any major trends down core.

Carbon data (Appendix 3) for both total and inorganic carbon show distinct differences between pink layers, silt layers, and background sediments (Figure 13). Pink silt clearly has the lowest total carbon values and the highest inorganic carbon values. This means that pink silt has, by far, the lowest organic carbon values. Brown silt has organic carbon values slightly lower than background and inorganic carbon values slightly higher than background, but clearly lower than pink silt (Figure 13). In profile, total carbon is spiky in nature and shows a drastic increase in the top 100 cm of Core A (Figure 12). Magnetic susceptibility through Core A(Appendix 4) is also spiky in nature showing a drastic decrease within the top 100 cm of the core and a slight increase past ~500 cm down profile (Figure 12). Finally, because bulk density (Appendix 5) reflects the amount of clastics compared to organics in a sample, as does, magnetic susceptibility, the two have a similar trend (Figure 12). There is slight offset between obvious correlating peaks in the two curves, but because magnetic susceptibility was measured before much shrinkage could have occurred, it is probably the more accurate depth. Radiocarbon dates from Core A and Core 2 (Appendix 6) indicate an age reversal in Core A and possible age reversals between Core A and Core 2 (Figure 14). The basal age of Core A indicated by radiocarbon is ~1000 yr.BP and the basal age of Core 2 indicated by radiocarbon is ~1500 yr.BP.


Brown Silt and Pink Silt

Brown and pink silt layers are clearly the two most distinctive and most abundant lamination types within Core A. Preservation of these layers in a currently eutrophic lake indicates that their deposition was either too rapid for bioturbation of the sediment to keep up or came at time when bioturbation was lower than present. In either case, because the layers are random, non-seasonal, and compositionally consistent down core, they probably represent instantaneous injections of allogenic sediment that have occurred by similar processes throughout the history of the lake. In the scope of the lifetime of a lake, the study area is virtually inactive in terms of earthquakes, which rules out tectonically induced sediment influxes. This leaves highly accelerated runoff during storms and flooding on the Mohawk River as the only possible mechanisms for creation of such distinct laminations in Core A.

The fact that the pink layers and brown layers are distinct from one another in terms of color, grain size, and carbon content indicates that they probably enter the lake by different processes and/or enter by the same processes but have different provenance. In general, brown layers are coarse (Figure 6), fairly organic rich (Figure 13), and coarsen upward (Figure 8) while pink layers are not quite as coarse (Figure 6), but are more clastic in composition (Figure 13) and fine upward (Figure 9) (Table 1).

Background Pink Silt Brown Silt

Coarsest Grain Size Fine-Silt Medium- to Coarse-Silt Coarse-Silt to Fine-Sand
Grain Size Trends Fine upward/smooth bell Coarsen upward/bimodal
Organic Carbon (avg.) 3.0 % 2.5 % 1.2 %
Inorganic Carbon 0 % - 0.23 % 0.11 % - 0.38 % 0 % - 0.36 %
Other Characteristics Erosive basal contacts

In the pink laminae, the fining-upward sequences and even grain size distributions without any basal scouring indicate that they are probably settling out of the water column in accordance with StokeÕs Law. The most probable mechanism for this type of deposition without basal scour is overflow, where the incoming waters, though filled with sediment, are not quite as dense as the lake waters. Additionally, the uppermost samples from two of the pink layers show the distinct shoulder characteristic to background sediments (Figure 9). The shoulder can be interpreted as the reintroduction of background influence during a waning event where rain out of allogenic sediment continues but is tinged by authigenic sediment.

Given that sediment within the drainage basin of Collins Lake is nearly all quartz sand, the abundance of pink minerals that give the pink layers their color indicates that their sediments must have come from outside the drainage basin. It follows then, that sediments making up the pink layers must have come from the suspended load of the neighboring Mohawk River during high flood stages. The Catskill Redbeds that are drained by Schoharie Creek, the main tributary to the Mohawk, provide a good source of pink clays. The twelve inches of rain in the Catskills during Hurricane Floyd produced distinctly red discharge in the Mohawk River (Garver, 2000, Personal Communication). Additionally, the relatively high inorganic content of the pink layers may be a product of the abundant carbonate rocks that are incised by the Mohawk River itself in the Mohawk Lowlands.

In contrast to pink layers, brown layers are coarser grained (Figure 6), more organic rich (Figure 13), coarsen upwards (Figure 8), and exhibit basal scouring (Figure 4) (Table 1). Though difficult to explain, the coarsening upward sequences eliminate overflows as a mechanism of deposition, and basal scouring alone indicates that the layers must have been deposited in underflows (Allison, 1995). Because of their relatively high organic carbon content and grain sizes reaching into the fine sand range, it is unlikely that sediment making up the brown silt layers came from the suspended load of the Mohawk River. The cliffs of the drainage basin, however, could provide more than enough fine sand to make up the grain size distribution of the brown layers, and their leaf litter could contribute to the fairly high organic component of the brown layers. Additionally, the high gradient of the slope may provide adequate means for producing sediment-laden underflows. It may be the case, however, that the brown silt layers are merely the result of higher magnitude flooding, allowing bedload from the Mohawk to get into Collins Lake.

From 350 to 500 cm below the water/sediment interface there are frequent pairs pink silt and brown silt layers where pink silt layers directly overlay brown silt layers. Grain size trends from within these pairs (Figure 10) illustrate that the pink mud contains a coarse signal from the underlying silt layer, which is not seen in any other pink layers. Because the pair is intact and shows no signs of bioturbation, the coarse signal in the pink layer indicates that the two were deposited as part of the same event. If the brown silt is indeed bedload from the Mohawk and the pink silt is suspended load, then a pink silt layer raining out on top of a previously injected brown layer during a waning flood event makes sense. If this were the case, however, a solitary brown layer would not be expected at any depth within the core because any injection of bedload would certainly bring in suspended load as well. Since there are plenty of solitary brown layers in Core A, (Figure 4), a different scenario is suggested. If the brown silt layers were coming from the hillslopes of the drainage basin, then during a large-scale storm, the sediment on the hillslope gets mobilized first, followed by spilling of the Mohawk into Collins because of the lag time in river response to storms. Thus, a solitary brown layer could be evidence of a storm that was big enough to mobilize sediment on the hillslope, but not big enough get floodwaters from the Mohawk into the Lake.

Down Profile Variations

Radiocarbon dates from Core A indicate a basal age of ~1000 yr.BP (Figure 14). Given that there are age reversal in the core, age constraints are maximums at best, so sedimentation rates are assumed to be constant through time, yielding a sedimentation rate of 0.57 cm/year for Core A. The radiocarbon age of 710 +/- 50 yr.BP obtained from 248-252 cm below the water/sediment interface in Allison 1995 was transferred to Core A using major spikes in magnetic susceptibility that match strongly between the two cores, (Figure 15). This date indicates a sedimentation rate of 0.63 cm/year for Core A. Thus, a sedimentation rate of 0.6 cm/year was adopted for all of Core A.

There are three major changes in sediment parameters, ~75 cm, ~250 cm, and ~520 cm, occurring within Core A. The first, ~520 cm, occurred around 1150 AD, one hundred years after the birth of the lake where consistently high bulk density and magnetic susceptibility values decreased slightly and became more spiky in nature (Figure 12). Additionally, the grain size is generally coarser and pink laminations are much more frequent below 520 cm (Figure 4). These trends can be attributed to a post avulsion scenario where the young lake was still very intimately connected to the Mohawk River and received flood sediments regularly, even under very low flood stages. It is also important to note that brown silt layers were not more abundant, in fact they may have been less abundant during this time, which further supports the argument that pink layers are Mohawk derived while brown layers are from within the basin. If brown layers were Mohawk derived then one would expect them to be more abundant in this section of the core as well.

From ~520 to ~250 cm the core contains abundant pink and brown silt layers (Figure 4) and has correspondingly variable bulk density, magnetic susceptibility, and total carbon values (Figure 12). This section also contains the majority of the sand and coarse silt spikes in the core (Figure 12). The abundance of pink silt laminae indicates that Mohawk flooding was either more frequent or could reach the lake more easily during this time, and the abundance of silt layers indicates that climate must have been stormier to have mobilized more sediment off the slopes surrounding the lake. The change in both of these parameters, of course, could be due to a change in productivity where bioturbation increased, but since the change occurred ~1590 AD, well before any foreign species affected the lake, it is likely a true climate signal. Additionally, the 1590 AD change occurred before humans would have had any effect on the system, which leaves climate as the likely driver.

Grove (1988) identifies a period from 1350-1850 called the Little Ice Age where temperatures in Europe were ~1? C colder than today, a change that could have caused increased storminess leading to the abundant pink and brown silt layers in the bottom half of Core A. The signal in Core A, however, spans from 1180 AD to 1600 AD, about two hundred years too early for the Little Ice Age. Powell (1992) recognizes that the Little Ice Age varied both in space and time, so the Collins Lake period of increased storminess may have indeed been caused by the Little Ice Age. The lack of a strong age chronology to solidify changes in sedimentation rates with time may also lead to error causing the 200-year offset. At any rate, because the pink and brown layers vary together in this section of the core, the change was probably induced by one allogenic process, leaving climate as the likely option.

The final change in core parameters occurred at ~75 cm, which correlates to the year 1875. The uppermost 75 cm of Core A is characterized by a sharp increase in total carbon and a subsequent decrease in MS and bulk density (Figure 12). The sediment is also massive and featureless in this section (Figure 4). This change is well within the limits of the introduction of foreign plant species into the lake, which contributed to its eutrophication. This increased productivity could well have enhanced bioturbation enough to mix any laminae that may have come into the lake. It also could have provided enough authogenic carbon to contribute to the swing in the carbon curve. It is likely, however, that a combination of effects was at work. The building of dikes around the lake in 1804, 1835, and 1946 would have inhibited Mohawk flood-waters from entering the lake and diluting the authogenic carbon thus allowing the carbon to increase like it does.

Also of interest in the profile of Core A is the drastic drop in total cabon and increase in inorganic carbon at 7 cm. The carbon signal is convincingly similar to a pink silt layer and may correspond to the 6 m flood on the Mohawk in 1996, the only high magnitude Mohawk flood in the past 60 years (Scheller, 2000). At a sedimentation rate of 0.6 cm/ year, the spike corresponds to the year 1989, an error that could easily be accounted for in an inaccurate C14 date or an erroneously low sedimentation rate.

Finally, an extended drop in total carbon around 1875 and another drop in total carbon around 1920 correspond with a period of abundant high magnitude floods on the Mohawk River (Scheller, 2000). This period also contains the only spike in sand in the top half of the core (Figure 4). A wet period in the northeast from 1840 to 1915 is recorded in Chesapeake Bay sediments (Cronin, et. al., 2000), which may be the driver behind the storm events recorded in carbon values during this time.

Conclusions and Discussion

Pink silt layers in Collins Lake stratigraphy represent overflows spilling in from the Mohawk River during major flood events while brown silt layers represent underflows of sediment coming off the banks of the drainage basin during large-scale storms. Covariations of these layers indicate a time of increased storminess from 1180 to 1600 that may be the result of the Little Ice Age. Near the top of the core, eutrophication by foreign plant species and/or construction of dikes isolating the lake caused a drop in lamination preservation, perhaps masking a second wet period from 1840 to 1915.

The lack of good age constraints and the assumption that sedimentation rates have been constant through time limits the interpretations of this study markedly. The use of grain size, carbon content, and color as provenance indicators is also limited. Using a better signature, perhaps rare earth elements, to distinguish the two would be useful.


First and foremost, I would like to thank Don Rodbell, for all of his excitement, knowledge, and advice, it is greatly appreciated. Thanks also to John Garver for his thoughtful attention, Jamie Garrand for her core lab expertise, and Carl George whose papers are referenced way too many time in my thesis. I extend special thanks to Jason Lederer, the authority on Mohawk River flooding. And finally, to give credit where credit is due, I thank my parents.


Allison, K., 1995, The sedimentologic history of Collins Lake, Scotia, NY, Senior Thesis, Union College Geology Department.

Bierman, P., Lini, A., Zehfuss, P., Church, A., Davis, P. T., Southon, J., Baldwin, L., 1997, Postglacial ponds and alluvial fans: recorders of landscape history: GSA Today, v. 7, p. 1-8.

Caldwell, D., and FitzGerald, D., 1995, Origin of lake-outlet deltas in Maine: Sedimentary Geology, v. 99, p. 95-110.

Cronin, T., Willard, D., Karlsen, A., Ishman, S., Verardo, S., McGeehin, J., Kerhin, R., Holmes, C., Colman, C., Zimmerman, A., 2000, Climatic variability in the eastern United States from Chesapeake Bay sediments: Geology, v. 28, p. 3-6.

Dlubac, J., 1995, Fossilized pollen grain analysis and downcore determination of exchangeable metal concentrations in Collins Lake, Senior Thesis, Union College Geology Department.

Gara, J., 1999, Living on the Mohawk River: a chronology of the geology, glaciers, bridges and history that shaped a river, Masters Thesis, Union College Geology Department.

George, C., 1978, The origin and early history of Collins Lake.

George, C., Tobiessen, P., Snow, P., Jewell, T., 1982, The monitoring of the restorational dredging of Collins Lake, EPA Project Summary, Scotia, NY.

George, C., Joachim, A., 1993, Long-burried lead shot: stability, possible transport by waterfowl, and re-exposure by hydraulic dredging, Collins Lake, Schenectady Co., New York, New York State DEC.

George, C., 1999, A limnologic chronology for Collins Lake and vicinity, Scotia Park Board.

Grove, J., 1988, The Little Ice Age: New York, Methuen, 498 p.

Howk, S., 1995, A record of the frequency of fires as derived from the abundance of charcoal in a sediment core from Collins Lake, Scotia, NY, Senior Thesis, Union College Geology Department.

Lederer, J., Garver, J., 1999, Ice jams during the 1996 mid-winter flood on the Mohawk River, New York: a reconstruction of ice jam locations based on ice scarring to bank-margin trees, Research Project, Union College Geology Department.

Luey, K., 1999, A chronology of major floods on the Mohawk River in Schenectady: 1832 to 1903, Research Project, Union College Geology Department.

Powell, C., 1992, A confusing Little Ice Age: Scientific American, v. 266, p. 21.

Rasanen, M., Salo, J., Jungner, H., 1991, Holocene floodplain lake sediments in the Amazon: 14C dating and palaeoecological use: Quaternary Science Review, v. 10, p. 363-372.

Ruggiero, K., 1999, The geological evolution of Collins Lake, Scotia, New York, as revealed from sub-bottom profiles and sediment core analysis, Senior Thesis, Union College Geology Department.

Scheller, M., 2000, A chronology of flooding on the Mohawk River, Senior Project, Union College Geology Department.

Van der Bogert, M., 1978, A chronology of the floods and flood related events in the Schenectady section of the Mohawk River, Schenectady Chamber of Commerce Flood Assessment Task Force, 14 p.


Funded, in part, by the National Science Foundation

© Geology Department, Union College, Schenectady N.Y. 12308-3107.

© Environmental Science and Policy Program, Union College, Schenectady N.Y. 12308-3107. All rights reserved. No part of the document can be copied and/or redistributed, electronically or otherwise, without written permission from the Director of the Environmental Studies Program, Union College, Schenectady NY, 12308-2311, USA.

This document can be located from http://minerva.union.edu/garverj/mohawk