Coevolution of Flood Dynamics and Economical Production in Tropical Wetlands: Insights From Bayesian Networks in Ayapel Wetland, Colombia

Abstract

AbstractThe coevolution approach in socio‐hydrological systems (SHS) allows the analysis of the interactions between social and hydrological systems over time. SHS's characteristics such as non‐linearity, causality, integration of spatiotemporal scales and feedbacks must be considered. However, in simulation processes, there are challenges in analyzing these characteristics while integrating quantitative and qualitative source variables. We propose a general causal network for study of coevolution in tropical wetlands, operationalized through Bayesian Networks for a case study in Ayapel Wetland, Colombia. We defined the flood pulse and economic production to analyze the co‐evolution of tropical wetlands. We identified that the probability of having more than 146,998 individuals of livestock is 30% and of obtaining a fish production considered “high” by fishers is 44.2% for a period from 1995 to 2019. However, under permanent conditions, the probability of this livestock decreases to 18.7%, and “high” fishing increases to 87.5%. Under seasonal conditions, the probability of high livestock production increases to 42.8%, and fish production decreases to 3.85%. We demonstrated how the loss of functionality of the existing hydraulic technology in the study area modifies the production dynamics and how a maximum level of agricultural and livestock production is associated with an increase in forest and grassland cover. We consider that the application of the network strengthens adaptation strategies for climate change or The El Niño‐Southern Oscillation events, the risk management plans and land use decision plans in the tropical wetlands.