Repository files navigation

Mizer is an R package to rundynamic multi-species size-spectrummodels of fishcommunities. The package has been developed to model marine ecosystemsthat are subject to fishing. However, it may also be appropriate forother aquatic ecosystems. By providing a framework for multi-speciesfisheries modelling as an R package, mizer enhances the accessibility,usability, and reproducibility of models, and thus aims to facilitatecollaboration and innovation.

The package contains functions that allow you to set up an ecosystemmodel and then project it through time under different fishingstrategies. Methods are included to explore the results, including plotsand calculations of community indicators such as the slope of the sizespectrum. Size-based models can be complicated, so mizer contains manydefault options that you can however change when needed.

Mizer has been supporting research in marine ecology and fisheriesscience since 2014 (seepublications).Mizer is still under active development. Version 2.0 has increased theuser-friendliness and the flexibility of the framework. Contributionsfrom the user community are very welcome. There is a sister packagecalledmizerExperimentalwhere user contributions can be checked out and receive feedback fromthe community. Example mizer models can be contributed tomizerExamples. Follow us ontwitter and read ourblog to stay up-to-date with newdevelopments. Does your project or publication use mizer? If so, wewould love to know.

Recent work on mizer was funded by the European Commission Horizon 2020Research and Innovation Programme under Grant Agreement No 634495 forthe project MINOUW (http://minouw-project.eu/) and the AustralianResearch Council Discovery ProjectRewiring Marine FoodWebs.

The package is onCRAN andtherefore available from R’s built-in package manager.

# Install latest released version from CRANinstall.packages("mizer")# Alternatively, install the development version from GitHubremotes::install_github("sizespectrum/mizer")

The following code loads the mizer package, loads some information aboutspecies in the North Sea that comes as an example with the package, setsup the parameters for the mizer model, and runs a simulation for 10years.

library(mizer)params<- newMultispeciesParams(NS_species_params,NS_interaction)sim<- project(params,t_max=10,effort=0)

The results of the simulation can then be analysed, for example viaplots:

plot(sim)

See the accompanyingGetstarted page formore details on how the package works, including detailed examples.

Size-based multi-species models are important for fisheries sciencebecause they provide a more realistic and accurate representation of thedynamics of fish populations and the ecosystems in which they live. Incontrast to traditional single-species models, which consider a singlefish stock as an isolated unit, size-based multi-species models accountfor the fact that fish populations are part of a larger ecosystem andinteract with other species through predation, competition, and otherecological processes.

One of the key advantages of size-based multi-species models is thatthey provide a more comprehensive understanding of the impacts offishing on fish populations and ecosystems. By considering the sizedistribution of different fish species, these models can capture theeffects of fishing on both target and non-target species, and ondifferent life stages of a species. This is particularly important forspecies that are caught as bycatch or that are indirectly affected byfishing through changes in their food web.

Another advantage of size-based multi-species models is that they can beused to investigate the effects of environmental changes and otherperturbations on fish populations and ecosystems. For example, thesemodels can be used to explore the impacts of climate change on thedistribution and abundance of fish populations, or the effects ofhabitat loss or pollution on fish communities. Because mizer is amechanistic model, it can deduce the complex population-level changesthat we are interested in from the simpler changes in the physiologicalrates and feeding interactions of individual fish species.

Overall, size-based multi-species models provide a more comprehensiveand realistic framework for understanding the dynamics of fishpopulations and ecosystems, and for developing effective fisheriesmanagement strategies that account for the complex interactions amongspecies and their environment.

A mizer model captures the interactions between species. The growthrates of fish are determined by the availability of prey and the deathrates are influenced by the abundance of predators, as well as fishing.The model starts with the individual-level physiological rates for eachspecies, as well as the predation preferences, and deduces thepopulation-level dynamics from these. Thus quantities like fish dietsand fisheries yields emerge dynamically and can be projected into thefuture.

Because a mizer model tracks the size of individuals as they grow upover several orders of magnitude from their egg size to their maximumsize, it correctly tracks the ontogenetic diet shifts. An individualtypically moves through several trophic levels during its life time.This is often not correctly captured in other multi-species models.

A mizer model can be set up with only a small amount of informationbecause it uses allometric scaling relations and size-based feedingrules to choose sensible defaults for unknown parameters.

Setting up a new multi-species mizer model is a two-step process,similar to what may be familiar from Ecopath with Ecosim: First onecalibrates the model to describe a steady state that is in agreementwith current observations (as in Ecopath), then one chooses theadditional parameters that determine the dynamics away from the steadystate (as in Ecosim). This model can then be used to investigate futureeffects of changes in fishing policy or of environmental stressors.

One big advantage of a mizer model is that it is based on a strongmathematical foundation. This allows a degree of a priori understandingof the behaviour of the model that is absent in many other multi-speciesmodels. This theoretical foundation is well presented in the book “FishEcology, Evolution, and Exploitation” by Ken Andersen.

It is interesting to think of the marine ecosystem as a transport systemthat moves biomass from the size of primary producers (mostlyunicellular plankton) up to the sizes of fish that humans like toconsume. Each fish that grows up from egg size to maturity by eatingsmaller individuals is like a car on this biomass highway. The yield ofour fisheries depend on this traffic flowing smoothly and withouttraffic jams.

An analogy with road traffic may be helpful:

In road traffic, if traffic density gets too high in a section of thehighway, drivers slow down, which leads to a pile-up producing evenhigher traffic density, leading to further slow-down in a potentiallyvicious cycle known as a traffic jam. Traffic management that ignoreshow the traffic density affects traffic speed fails. Luckily ourmathematical understanding of transport equations has made practicalcontributions to managing traffic in ways that produce smoother trafficflow and hence higher throughput.

Mizer implements the transport equations for marine ecosystems. Thepotential for traffic jams is the same: if for example there is a highdensity of predators of a particular size, which all have preference forprey of a particular smaller size, then due to competition for that preythe growth of those predators slows down, leading to a pile-up whichleads to further depletion of prey, leading to further slow-down, in apotentially vicious cycle.

Luckily, the natural ecosystem has evolved to facilitate very smoothtraffic on this biomass highway, with resultant high productivity. Thisstate is characterised by an approximate power-law shape of the biomasssize spectrum. The purpose of mizer is to allow us to understand howvarious stressors, like fishing or climate change, affect the sizespectrum and hence the flow of biomass and the productivity andresilience of the marine ecosystem. Mizer allows us to investigate howsize-based fisheries management strategies can be used to keep theecosystem close to its natural productive state.