Investigating ecosystem connections in the shelf sea environment using complex networks

.Weusecomplexnetworktheorytobetter representandunderstandtheecosystemconnectivityinashelfsea environment. Thebaselinedataused for the analysis are obtained froma state-of-the-art coupledmarine physicsbiogeochemistrymodelsimulatingtheNorthWestEuropean Shelf (NWES).Thecomplexnetworkbuilt onmodel outputsisusedtoidentifythefunctionalgroupsofvariablesbehindthebiogeochemistrydynamics,suggestinghowtosimplifyourunderstandingof thecomplexwebof interactions withintheshelf seaecosystem.Wedemonstrate that complexnetworkscanalsobeusedtounderstandspatialecosystemconnectivity, identifyingboththe(geographicallyvarying)connectivitylength-scalesandtheclustersofspatiallocations thatareconnected.Weshowthat thebiogeochemicallength-scalesvarysignificantlybetweenvariablesandare not directly transferable.Wealsofind that thespatial patternoflength-scalesissimilaracrosseachvariable,aslong asaspecificscalingfactor for eachvariable is taken into account.Theclusters indicategeographical regionswithin which there isa largeexchangeof informationwithin the ecosystem,while informationexchangeacross theboundariesbetween these regions is limited.The resultsof this studydescribe howinformation is expected topropagate throughtheshelfseaecosystem,andhowitcanbeusedin multiple futureapplications suchas stochasticnoisemodelling,dataassimilation,ormachinelearning.