PARENT SESSION
Poster Session 3: Aquatic Ecology
Monday, August 8, 5:00 PM - 6:30 PM, Exhibit Hall 210 D, Level 2, Palais des congrès de Montréal
A Spatially-Explicit Model of Iron Loading to Lakes.
Maranger, Roxane*,1, 2, Canham, Charles2, Pace, Michael2, Papaik, Michael2, 1 Université de Montréal, Montréal, Québec, Canada2 Institute of Ecosystem Studies, Millbrook, New York, US
ABSTRACT- We developed a general mass balance model of iron concentrations in lakes based on loading, within-lake processes and losses. We used data on watershed land cover, water color, and depth for 93 lakes in the Adirondacks of New York and maximum likelihood techniques to test alternative models. Loading to each lake was estimated as a function of vegetation cover type from 10 × 10 m grid cells within the watersheds and the flow path distance to the lake. In-lake production and losses of iron were estimated as a function of lake area and discharge. There was differential Fe loading from among the 8 vegetation types considered in this study, varying from negligible to 54 kg Fe ha-1 yr-1. Interestingly, the spatially explicit model that allowed Fe export to change as a function of distance to the lake was a poorer fit to the data when compared to the non-spatial model suggesting no distance decay. Areal export of Fe was greater from wetlands but wetlands on average only occupied 13% of watershed area. Hence, 75% of the Fe load was derived from upland forests. Fe losses were primarily regulated by discharge. However an alternative model, accounting for within-lake losses related to lake water color and depth was a better fit to the data thus providing more information. Overall, our results indicate that lake iron concentrations are determined by the balance between differential iron loading based on watershed land cover, lake discharge rates, and within-lake iron losses related to color and depth.
Key words: land-water interactions, Fe loading to lakes, wetlands, Maximum Likelihood