Modelling to inform the conservation and management of aquatic ecosystems: A synthesis of five case studies

The sustainability of aquatic ecosystems faces numerous threats (e.g. Dudgeon et al., 2006, Reid et al., 2019) that can challenge the management and sustainable extraction of resources from these ecosystems. Included among these challenges are human-driven environmental changes such as habitat degradation, aquatic invasive species (AIS), pollution (e.g. contaminants, nutrients, microplastics), over-exploitation, flow modifications, and climate change. These challenges are exacerbated by the inherent complexity of aquatic ecosystems and the potential for multiple stressors arising from human-driven environmental change to interact in novel ways (e.g. Scheffer et al., 2001, Brook et al., 2008, Côté et al., 2016). But the more interactions and inherent complexity, the more difficult it becomes to fully understand an ecosystem and how it may respond to management actions.An ecosystem approach to management can be used to consider a range of factors in management decisions (e.g. Garcia et al., 2003), and models provide tools that can help inform the management of complex systems. A model is simply a framework for synthesising information to improve understanding and decisions. Explicit models, whether they are formal mathematical models or conceptual models, provide structures that can be examined and used to make predictions. Despite their potential benefits, there is a false perception that models are rarely used to inform management (Schuwirth et al., 2019, DeAngelis et al., 2021). For models to be useful for management, they need to address the issues relevant to the decisions faced by managers. To accomplish this, three things need to be considered when choosing a modelling approach to inform management: (1) the management objectives, (2) the model outputs required to inform the management objectives, and (3) the data available to address the management objectives.This special issue provided five case studies (Table 1) of the application of modelling to management issues in aquatic ecosystems, including fisheries (Weyl et al., 2023), aquaculture (Muralidhar et al., 2023), freshwater availability (Das et al., 2023), water quality (Zhang et al., 2023), and foodweb responses (Giacomini et al., 2023; Zhang et al., 2023). These papers do not rely on a single modelling approach but provide examples of the application of models that are tailored to the needs of the management issues of concern with the available data. These case studies range from data limited to data rich applications, making use of surplus production and statistical catch-at-age models (Weyl et al., 2023), system dynamics modelling (Muralidhar et al., 2023), downscaled global climate modelling (Das et al., 2023), size spectra (Giacomini et al., 2023), and Ecopath with Ecosim (Zhang et al., 2023).Muralidhar et al., (2023) provided a review of models related to the complex interactions between environmental factors such as temperature, salinity and dissolved oxygen, and shrimp growth, survival, and disease incidence in aquaculture facilities. The aim of this review was to understand the interactions and feedback for integrated coastal aquaculture management and planning to improve the sustainability and productivity of shrimp aquaculture systems. Muralidhar et al., concluded that system dynamics modelling can provide a useful tool for understanding and optimizing coastal aquaculture systems that are complex and involve a wide range of interconnected processes; they provided an overview of the strengths and limitations of some of the software that can be used to build system dynamics models. Muralidhar et al., also concluded that system dynamics models can help aquaculture managers make more informed decisions that lead to more efficient and sustainable shrimp production by considering the complex interactions among system components.Das et al., (2023) aimed to estimate freshwater availability in the Sundarbans region (India) under climate change scenarios to inform fisheries and aquaculture management. Their analysis of long-term trends (1948-2010) showed decreasing freshwater availability (rainfall minus evapotranspiration) during the monsoon season, with only one of 19 blocks examined showing an increase. Das et al., downscaled a global climate model (GCM) to improve climate projections over the Sundarbans region, better representing past trends than the raw GCMs. The downscaled GCM projected increased freshwater availability in the Indian Sundarbans under warming scenarios due to increased rainfall during the monsoons. While this may be beneficial to fisheries, heavy rainfall has the potential to damage infrastructure. However, there are many uncertainties and Das et al., recommended that policymakers prepare for the worst-case scenario when formulating adaptation strategies.Weyl et al., (2023) used a multiple model approach to assess the current status of seven fish species encountered by the commercial trawl fishery in the southern part of Lake Malawi (Africa). This task was challenging given the data limitations and the need to organize a time series of data on these species. Weyl et al., were able to apply surplus production and statistical catch-at-age (SCAA) models to show that three of the fish stocks were within sustainable limits, four stocks appeared overfished in recent years, though one stock may have recently recovered. This data-limited modelling application demonstrated that modelling frameworks can be applied in data-deficient fisheries. This application also demonstrated that model outputs that can inform management are possible with a base level of data, leading Weyl et al., to recommend that a systematic monitoring and evaluation program, like the one being conducted in Lake Malawi, should be developed for other inland lake systems.Zhang et al., (2023) used Ecopath with Ecosim (EwE) to investigate the separate and interactive effects of nutrients and hypoxia on the foodweb structure of the central basin of Lake Erie (North America) to inform nutrient management strategies. Their model predicts that nutrients have a greater effect on the foodweb than hypoxia, with positive, non-linear effects on the biomasses of all foodweb components. Hypoxia, however, had mixed effects resulting in the decreased biomasses of benthic invertebrates, benthivorous fishes, and some omnivores but increased the biomasses of plankton and planktivorous fishes. Additionally, there were interactive effects of hypoxia and nutrients. Hypoxia amplified the effects of nutrients on pelagic foodweb groups and depressed the effects of nutrients on demersal foodweb groups. This application demonstrated the potential of foodweb models to incorporate multiple stressors to better understand how they may interact and affect the anticipated outcomes from management actions.Sometimes modelling approaches need to be further developed to serve management needs. Giacomini et al., (2023) did this with size spectra. The management of inland lakes in Quetico Provincial Park (Ontario, North America) is based on indicators to evaluate the biological status of lake communities. The slope of the size spectrum can provide an indicator of ecosystem processes and responses to perturbations. Giacomini et al., explored an alternative approach to fitting the size spectrum to account for the use of multiple sampling protocols and distortions caused by gear size selectivity. They found that this analytical method can integrate body size data from multiple taxonomic or functional groups and sampling protocols, while combining the flexibility of linear regression with the statistical rigor of distributional approaches. Their analyses also showed that a wide range of trophic groups is essential for the size spectrum to be useful as a consistent indicator of energy transfer. In Giacomini et al., development of the fitting approach to size spectra demonstrated that consideration of the specifics (and potentially limitations) of the available data can lead to improvements in both the modelling approach and the information provided to management.While this special issue is not exhaustive, it provides a cross-section of the diverse types of modelling approaches available and how they can be used to inform management. These case studies come from multiple types of aquatic ecosystems, across three continents, addressing different management issues and objectives. What they have in common is the demonstrated application of modelling approaches, chosen specifically to address the management needs for the ecosystem under consideration within the limitations of the available data. By providing examples of models applied to management needs, these studies point the way toward increased use of these synthesising tools to expand the use of models to inform ecosystem management decisions.