| Paleolimnological reconstructions have demonstrated an inherent variability in phosphorus concentrations among lakes prior to anthropogenic landscape disturbances. However, the natural drivers of this variability are less frequently studied than agricultural or urban phosphorus loading to freshwaters, and are not as well understood. Therefore, this research examined pre-settlement phosphorus concentrations of 48 inland lakes in Michigan in conjunction with pre-settlement land cover data, hydrogeomorphic features, and geographic variables, to better understand these natural landscape drivers. This multi-step process involved an analysis of diatom assemblages from lacustrine sediment cores, relative dating of core samples using Ambrosia pollen, a native land cover analysis in GIS, and finally, regression modeling of the data. The results of OLS regression modelling suggest that the variability in pre-settlement phosphorus concentrations of my dataset is best explained by native wetland cover. Moreover, native wetlands at a local (100 m buffer) scale had a stronger relationship with pre-settlement phosphorus concentrations than native wetlands within the entire lake catchment. The strength of this relationship improved further when wetlands were modeled with maximum lake depth. My analyses show that as the amount of pre-settlement phosphorus concentrations in a lake increased as wetlands near the lake shore increased. Furthermore, as lake depth decreases, the effect of wetlands on lake phosphorus concentrations increases. However, because my study was only able to explain 34% of the variability in pre-settlement phosphorus concentrations, further research is warranted to better understand this natural variability. |
