Abstract
Input–output nutrient budgets were constructed for 8 and 10 diked contiguous 4- to 6-ha marshes in 1981 and 1982, respectively. Because the marshes studied were created by diking within a larger marsh, the water budgets were dominated by water pumped in to replace roughly equal evapotranspiration and seepage losses. Chloride budgets could not be used to confirm water budgets, apparently because of a transfer to the sediments by a freezing-out process in winter and a subsequent diffusion back into the surface water during the ice-free season. With replicated ponds and accurate water budgets, good agreement was found among replicates for annual budgets for the major ions and total N. However, the budgets for total P were highly variable, reflecting the biological activity of P. Atmospheric inputs of N and P were large components of budgets for these nutrients. Year to year increases in surface water concentrations of major ions, expected on the basis of evaporative water loss, did not occur, apparently because of freezing-out of these ions in winter.