A Holocene record of endogenic iron and manganese precipitation, isotopic composition of endogenic carbonate, and vegetation history in a lake-fen complex in northwestern Minnesota

Abstract

Little Shingobee Lake and Fen are part of an extensive network of lakes and wetlands in the Shingobee River headwaters area of northwestern Minnesota. Prior to about 9800 radiocarbon years, most of the lakes in the Shingobee watershed area were interconnected to form glacial Lake Willobee. From 9800 to 7700 radiocarbon years, the level of Lake Willobee fell as a result of breaching of a dam, leaving small separated basins containing the existing lakes and wetlands. The dominant components in the sediments in a 9-meter core from Little Shingobee Lake (LSL-B), and lacustrine sediments under 3.3 meters of peat in a 17-meter core from Little Shingobee Fen (LSF-10) are detrital clastic material, endogenic CaCO 3 , and organic matter. The detrital fraction in the Holocene section in core LSL-B varies considerably from 7 weight percent to 82 weight percent and closely parallels the concentration of detrital quartz measured by X-ray diffraction. The CaCO 3 concentration, which also varies considerably from 10 weight percent to 70 weight percent, is generally antithetic to the detrital concentration owing to the dilution of detrital material by CaCO 3 , particularly during the early to middle Holocene (about 9000-6500 calendar years). The organic-matter content varies from 5 weight percent to 25 weight percent and, together with CaCO 3 , serves to dilute the allogenic detrital fraction. In both cores almost all of the iron (Fe) and manganese (Mn) is in endogenic minerals, presumed to be oxyhydroxide minerals, that are important components throughout the core; little Fe and Mn are contributed by detrital aluminosilicate minerals. The endogenic Fe mineral, calculated as Fe(OH) 3 , forms a larger percentage of the sediment than endogenic organic material throughout most of the Holocene section in the LSL-B core and in the lacustrine sediments below the peat in the LSF-10 core. Biogenic silica as opal (biopal; diatom debris) was not measured, but the average calculated biopal is 5 percent in the LSL-B core and 15.5 percent in the LSF-10 core. Values of delta 18 O in mollusk (Pisidium) and ostracode shells increase by only about 20 per mil from the bottom to the top of the LSL-B core (about 12600-2200 calendar years). The remarkably constant oxygen-isotope composition throughout the Holocene suggests that environmental conditions affecting values of delta 18 O (temperature, salinity, composition of the water, composition of precipitation) did not change greatly. Values of delta 13 C in carbonate shells generally decreased by about 2 per mil from 9000 calendar years to 6000 calendar years, but they did not increase in organic carbon. This mid-Holocene increase in delta 13 C in shells but not in organic carbon is likely due to an increase in residence time. A late Pleistocene forest dominated by spruce was replaced in the early Holocene by a pine forest. The pine forest migrated east during the middle Holocene and was replaced by an open sagebrush-oak savanna. The western migration of forests into northwestern Minnesota is marked first by a hardwood forest and finally a pine forest.