Quantifying the response of a functional indicator of ecosystem health to disturbance gradients in New Zealand riverine environments: a meta-analysis
In New Zealand, recent policy changes require freshwater managers to take more comprehensive and integrated approaches to monitoring and maintaining ecosystem health. To attempt to prevent and reverse the adverse effects of land use change on freshwater ecosystems, management decisions need to be based upon a suite of indicators each with a strong foundation of knowledge regarding the nature of responses at a national scale. Monitoring ecosystem function in addition to structural indicators has long been suggested to provide a more accurate and holistic narrative of ecosystem health, however, it has yet to be adopted in routine bioassessment. The cotton strip assay has shown promise as a consistent, relatively cheap, and repeatable method for monitoring freshwater ecosystem function, indicating the ecological processing rates of riverine microbial communities and the organic matter processing potential of riverine environments. Numerous regional-scale studies have applied the cotton strip assay in New Zealand, but these data have yet to be explored in unison. For managers to successfully monitor, manage, and restore ecological processes in river environments, a comprehensive understanding of the proximate drivers of cotton breakdown is needed. The aim of this study is to conduct a meta-analysis of cotton strip assay data to explore the relationship between river function and other measures of ecosystem health and land-use stressors at a national scale.
I collated published and unpublished cotton strip data to create a meta-dataset, with measures harmonised by deployment time and temperature for more meaningful comparisons at a national scale. I sourced additional data from national databases describing water quality and physical river classification information for more comprehensive, higher resolution analyses. I then used the meta-dataset was to investigate the nature of cotton decomposition responses along varying levels of impairment across different seasonal conditions and spatial catchment attributes.
I used linear mixed-effects models to determine the relationships between cotton decomposition and physicochemical predictor variables, along with any additional influence attributed to underlying spatial variation across sites. Results suggest that bioavailable nutrients and water clarity are the largest drivers in cotton breakdown rates at a national scale. Water temperature and seasonal conditions emerged as likely limiting factors on microbial activity and cotton breakdown, indicating that consistent intra-seasonal monitoring is advisable. Climate and underlying geology can also be important when looking to discriminate underlying catchment variation and should be incorporated when making larger scale comparisons. Relationships with land use were found to be non-linear and likely to have too many co-varying factors enacting influence on cotton breakdown rates to be successful predictive gradients. Breakdown responses were, however, most consistent under high levels of vegetation cover, and high variability in responses in more urban and pastoral developed catchments. The assays’ sensitivity to nutrient enrichment at a national scale could aid in informing management policies with respect to nutrient limits, and the setting of natural ecosystem processing benchmarks.