Scaling-up ecosystem functions of coastal heterogeneous sediments: testing practices using high resolution data

Abstract

Context: Heterogeneity in coastal soft sediments and the difficulty of data collection hinder our ability to scale up ecological data (necessarily obtained at small-scale) to large-scale. The use of scaling in marine ecology is not as common as in terrestrial ecology and current practices are often too simplistic and inadequate. Objectives: We aimed to demonstrate that the use of different scaling approaches leads to considerably different results and that not accounting for ecological heterogeneity decreases our ability to accurately extrapolate measurements of ecosystem functions performed by intertidal soft sediment habitats. Methods: High resolution raster maps of sediment denitrification, ammonia (NH4+) efflux and organic matter degradation were sampled to produce a simulated dataset and compare the performance of three different scaling approaches: direct scaling, spatial allometry and semivariogram/kriging. Results: Direct scaling underestimated denitrification, NH4+ efflux and organic matter degradation (84.1, 84.9 and 90.3% less) while allometry underestimated denitrification (81.9% less) but overestimated NH4+ efflux and organic matter degradation (2594.1 and 14,879.9% more). Kriging produced more accurate results and the predicted functions only differed from the estimated values by 14.7, 29.4 and 3.9% respectively. Conclusions: Our work shows that the choice of the scaling method is crucial in estimating intertidal soft sediment functions and highlights the need for empirical and theoretical models that link ecosystem functioning to biological attributes that can be measured remotely over large areas. Integrating measures of heterogeneity through the spatial structure of the data leads to outcomes that are more realistic and relevant to resource management.