Processes and predictions in ecological models: logic and causality

TY - JOUR

T1 - Processes and predictions in ecological models

T2 - logic and causality

AU - Damgaard, Christian

PY - 2025/8

Y1 - 2025/8

N2 - To make credible ecological predictions for terrestrial ecosystems in a changing environment and increase our understanding of ecological processes, we need plant ecological models that can be fitted to spatial and temporal ecological data. Such models need to be based on a sufficient understanding of ecological processes to make credible predictions and account for the different sources of uncertainty. Here, I argue (1) for the use of structural equation models in a hierarchical framework with latent variables and (2) to specify whether our current knowledge of relationships among state variables may be categorized primarily as logical (empirical) or causal. Such models will help us to make continuous progress in our understanding of and ability to predict the dynamics of terrestrial ecosystems and provide us with local predictions with a known degree of uncertainty that are useful for generating adaptive management plans. The hierarchical structural equation models I recommend are analogous to current general epistemological models of how knowledge is obtained.

AB - To make credible ecological predictions for terrestrial ecosystems in a changing environment and increase our understanding of ecological processes, we need plant ecological models that can be fitted to spatial and temporal ecological data. Such models need to be based on a sufficient understanding of ecological processes to make credible predictions and account for the different sources of uncertainty. Here, I argue (1) for the use of structural equation models in a hierarchical framework with latent variables and (2) to specify whether our current knowledge of relationships among state variables may be categorized primarily as logical (empirical) or causal. Such models will help us to make continuous progress in our understanding of and ability to predict the dynamics of terrestrial ecosystems and provide us with local predictions with a known degree of uncertainty that are useful for generating adaptive management plans. The hierarchical structural equation models I recommend are analogous to current general epistemological models of how knowledge is obtained.

KW - ecological prediction

KW - hierarchical model

KW - structural equation model

UR - https://www.scopus.com/pages/publications/85218775428

U2 - 10.1002/for.3267

DO - 10.1002/for.3267

M3 - Journal article

SN - 0277-6693

VL - 44

SP - 1658

EP - 1665

JO - Journal of Forecasting

JF - Journal of Forecasting

IS - 5

ER -