Integrating simulation models and statistical models using causal modelling principles to predict aquatic macroinvertebrate responses to climate change

Abstract

Climate change is projected to threaten ecological communities through changes in temperature, rainfall, runoff patterns, and mediated changes in other environmental variables. Their combined effects are difficult to comprehend without the mathematical machinery of causal modelling. Using piecewise structural equation modelling, we aim to predict the responses of aquatic macroinvertebrate total abundance and richness to disturbances generated by climate change. Our approach involves integrating an existing hydroclimate-salinity model for the Murray-Darling Basin, Australia, into our recently developed statistical models for macroinvertebrates using long-term monitoring data on macroinvertebrates, water quality, climate, and hydrology, spanning 2,300 km of the Murray River. Our exercise demonstrates the potential of causal modelling for integrating data and models from different sources. As such, optimal use of valuable existing data and merits of previously developed models in the field can be made for exploring the effects of future climate change and management interventions.