Diatom-based transfer functions for pH and total phosphorus in Vermont, USA lakes

Abstract

At a landscape scale, lakes function as early warning signs of ecological change in response to environmental stressors. Changes to the terrestrial landscape such as alterations in land-use or land-cover alter the quality and quantity of subsidies delivered to downstream lakes, resulting in cascading impacts on aquatic communities and ecosystem function. Together with climate change processes that alter subsidy delivery and processing rates, lakes are inextricably connected to their adjacent landscape, acting as dynamic integrators of auto- and allogenic stressors. These processes are recorded in lake sediments and can be inferred and reconstructed using paleolimnological proxies. Like much of the Northeastern USA, lakes in the state of Vermont are changing rapidly in response to multiple stressors. This includes more than 800 lakes that span gradients of elevation, latitude, trophic status, depth, clarity, and watershed area. Long term monitoring data indicates that oligotrophic lakes are experiencing increases in total phosphorus, and many montane lakes recovering from acidification are now experiencing browning and rapid surface water warming. Understanding these trends at long-term scales requires proxy-based reconstruction of sediment records. Here we describe regional training-sets and transfer functions for TP and pH developed using modern water chemistry and sediment diatom records from 80 and 96 lakes, respectively, spanning chemical, trophic, latitudinal, and elevation gradients. We reconstruct these variables for the period predating the European settlement of the state (~ 1800) using a top-bottom approach for 96 lake cores and present a high-resolution reconstruction of these variables for acid-impaired, mesotrophic Beaver Pond. Based on our findings, we present recommendations and limitations for model application.