CLIMATE CHANGE AND ECOSYSTEM RESPONSE AT THE BOUNDARIES OF THE MID-HOLOCENE DRY PERIOD BASED ON SEDIMENTS FROM TWO LAKES IN THE BIG WOODS OF MINNESOTA

Transitions from cool/humid to warm/dry (~8250 cal yr BP) and then to warm/humid (~4250 cal yr BP) conditions are characterized by major shifts in fire, vegetation, and lake ecosystems along the prairie-forest border in central North America.  We contrasted these two transitional periods (9000-7000 cal yr BP and 5000-3000 cal yr BP) based on high resolution (average interval 20-30 years) analysis of sieve charcoal, pollen, LOI, biogenic silica, and magnetics taken from the sediments of Sharkey Lake (44°35'39" N, 93°24'49" W), and Kimble Pond (44°13'15" N, 93°50'24" W) located in the Big Woods region of Minnesota.   During the older transition period decreases in ARM/IRM and increases in IRM at both sites indicate an initial rapid (~200 years) shift at ~8500 cal yr BP to warmer/drier climatic conditions followed by a continued but more gradual shift toward the generally warm/dry conditions characterizing the middle Holocene.  At Sharkey Lake vegetation and fire closely tracked the changes in magnetics with nearly 3-6 fold increases in grass pollen (~2% to 12%) and charcoal influx (~0.4 to 1.2 mm² cm^-2 yr^-1) occurring between 8400-8600 BP but this direct linkage was not observed at Kimble Pond, perhaps reflecting greater overall shift in vegetation at the former site. The shift to warm/humid conditions begins ~4100-4300 cal yr BP and also appears to have occurred in two steps but lake and landscape responses were generally more gradual and varied as exemplified by declines in biogenic silica, grass pollen and charcoal and a concomitant increase in Ulmus pollen at Kimble Pond but little or no change in pollen percentages, charcoal influx, or biogenic silica concentrations at Sharkey Lake during this time period despite marked changes in magnetics and LOI. Our data suggest that in Minnesota at the prairie-forest border that (a) changes in climate were greater and/or more rapid during the mid-Holocene shift to warm/dry conditions than the late-Holocene shift to warm/humid conditions and/or that (b) that responses of many common proxies to climate change as well as intercorrelations among proxies may be site specific suggesting that caution is necessary in developing regional models based on data from single lakes.