Scaling nested measurements of biogeochemical rates across prairie stream reaches with varying biotic and abiotic characteristics.

Printer-friendly version

Understanding the variability of ecological processes across spatial scales is a central issue in ecology, because increasing scale is often associated with increasing variability and complexity. Furthermore, ecological measurements are often made at scales smaller than those that management decisions are made. We generated a conceptual framework with predictions for scaling biogeochemical rates and tested our predictions by comparing measurements of ecosystem respiration (ER) and gross primary production (GPP) from patch to reach-scales (benthic areas of 1-100 cm2 or 10-100 m2 , respectively) over a two-year period across multiple sites (n= 7 for each rate). We evaluated interactive biotic and abiotic effects among patches by measuring biogeochemical rates in riffle and pool habitats with and without consumers, while quantifying a variety of abiotic variables (organic matter, water velocity, light availability, and substrata size) in each habitat. Consumer removal did not alter metabolic rates at the patch scale, suggesting abiotic conditions are more important drivers of these processes. Rate measurements in patches significantly overestimated reach ER and GPP after corrections for habitat heterogeneity, stream conditions (i.e. temperature and light) at the time of measurement, and abundance of alternative stream compartments (i.e. macrophyte beds and leaf packs). Our inability to scale is likely affected by different methodological approaches since reach scale measurements were in situ, while patch-scale measurements required removal of incubated substrata from the stream. Stream conditions were altered by drought and flood between years, and this translated into different up-scaled results, indicating that future climate scenarios should be considered when making predictions at any scale. While spatially explicit scaling approaches (i.e. mechanistic or process based modeling) may have been more effective for up-scaling metabolic rates in this system, this study provides valuable insight to factors that should be considered when attempting to scale ecosystem processes in any system.