Ecosystems

, Volume 15, Issue 3, pp 363–386

The Influence of Landscape Position and Catchment Characteristics on Aquatic Biogeochemistry in High-Elevation Lake-Chains

Authors

    • The Earth Research InstituteUniversity of California
  • Craig E. Nelson
    • The Earth Research InstituteUniversity of California
  • John M. Melack
    • The Earth Research InstituteUniversity of California
Article

DOI: 10.1007/s10021-011-9515-x

Cite this article as:
Sadro, S., Nelson, C.E. & Melack, J.M. Ecosystems (2012) 15: 363. doi:10.1007/s10021-011-9515-x

Abstract

To examine the influence of landscape characteristics and landscape position on aquatic biogeochemistry, we sampled a total of 76 lakes within 14 different lake-chains spanning the latitudinal extent of the high-elevation Sierra Nevada (California). We measured water chemistry, dissolved organic matter (DOM), nutrients, and biotic variables in study catchments that encompassed representative ranges of area (3–22 km2), elevation (2,200–3,700 m.a.s.l), elevation change (50–700 m), and average slope (13°–26°). Hierarchical models were used to account for variability in biogeochemistry because they explicitly maintain the framework of lakes within individual lake-chains while accounting for variation among lake-chains. Unconditional means models, where lake-chain was a random effect, revealed significant differences among lake-chains for nearly all biogeochemical variables. Models explained 42–95% of this variability, with the majority of the variation (70%) explained by the among lake-chain component. To explore the amount of additional variation explained by lake landscape position, we added lake network number (LNN) to models. LNN explained a significant amount of additional variation (7% average) in 8 of 23 biogeochemical parameters. However, it explained more variability within individual lake-chains (75%), where among lake-chain differences did not obscure patterns. Patterns of increase with LNN were found for dissolved organic carbon and nitrogen, fluorescence of DOM, alkalinity, and bacterioplankton abundance, whereas nitrate and nitrogen to phosphorus nutrient ratios decreased. LNN explained variation because it served as a proxy for underlying catchment characteristics that changed consistently along downstream flow paths. To characterize the amount of variation explained by catchment characteristics alone, we fit a third model that included lake-chain as a random effect and landscape or lake morphometry attributes as fixed effects. Catchment characteristics explained about as much additional variation (6%) as LNN, but for substantially more biogeochemical parameters (18 out of 23). The catchment characteristics most predictive of biogeochemistry were land-cover factors delineating alpine and subalpine zones (elevation, slope, or proportions of rock or shrub cover). In general, catchment characteristics were stronger predictors of biogeochemistry than characteristics of lake morphometry, emphasizing the relative importance of landscape processes in snowmelt-dominated lake ecosystems.

Keywords

landscape limnologylake-chainlandscape positionSierra Nevadahigh-elevationbiogeochemistrynutrient limitation

Copyright information

© Springer Science+Business Media, LLC 2011