Individual-based modeling of mangrove forest growth: MesoFON - Recent calibration and future direction

摘要

We introduce individual-based models (IBMs) of mangrove forests and criticize the tasks for their development recommended previously for being mostly related to natural threats. This is contrasted with our perspective that the key research question of today's models should be to mitigate anthropogenic threats.Core objective (1) of this article is to provide a review of mangrove threats prioritizing solution-oriented IBM approaches. Because species-specific calibration of IBMs is time-consuming, efficiency is crucial. Globally, we identify an urgent need to parametrize Asian mangrove species.We suggest IBMs to unveil management scenarios with maximum sustainable timber yield to prevent mangrove conversion and over-exploitation. The key model purpose regarding natural threats is to govern the management of mangrove forest stability for coastal protection using a combination of windthrow models and IBMs. We argue for the embedding of IBMs in ecosystem models to achieve purposes regarding eutrophication and altered hydrology/sedimentation.Core objective (2) is to describe the development of the new IBM mesoFON from a task-to a solution-oriented model. Initially, the interaction of lateral crown displacement and hurricane impacts was examined with mesoFON. Later, we introduced propagule production & local dispersal with the task to close the tree life cycle. Here, we describe the change in purpose of mesoFON accompanying its calibration for Rhizophora apiculata in Malaysia. For this we applied a Genetic Algorithm optimizer, used mesoFON as a "way-back machine", initialized it with observed tree diameters/positions and shrank the trees backwards in time.Objective(3) is to describe mesoFON's future direction: Embedding in the General Ecosystem Model (Fitz et al., 1996) and targeting the solution of threats at larger spatial scales. Finally, we demonstrate that the new model simulates overland waterflow qualitatively right even in benchmark settings.