ebook img

A model for an aquatic ecosystem PDF

0.09 MB·
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview A model for an aquatic ecosystem

A model for an aquatic ecosystem HanLiQiao,EzioVenturino DepartmentofMathematics“G.Peano”,UniversityofTorino,viaCarloAlberto10,I–10123Torino,Italy Abstract 7 An ecosystem made of nutrients, plants, detritus and dissolved oxygen is presented. Its equilibria are established. 1 0 Sufficient conditions for the existence of the coexistence equilibrium are derived and its feasibility is discussed in 2 everydetail. n Keywords: aquaticplants,dissolvedoxygen,nutrientrecycling a J 3 2 1. Introduction ] Eutrophicationis the large increase of aquatic plants due to the discharge in waterbodies of chemical elements E suchasnitrogenandphosphorus.Deadaquaticplantsgeneratedetritus,exploitinglargeamountofdissolvedoxygen P for its decomposition, and therefore endangeringaquatic life and fish survival. We introduce here a mathematical . o modelforsuchanecosystemandinvestigateitsbehavior. i b - q 2. Modelformulation [ LetN, DandOrespectivelydenotethenutrients,detritusanddissolvedoxygenamountinthewater. Further,let 1 P denote the aquatic plants population. The nutrients N are assumed to be dischargedinto the lake at rate q; they v alsoaregeneratedbythereconversionofdecomposeddetritus,atrateu,afraction0< f <1ofwhichbecomesnew 1 3 nutrient.Thisprocessinvolvesalsotheusageofdissolvedoxygen,andthereforethetermrepresentingitmustcontain 9 theproductofboththerequiredquantities,OandD. Inaddition,nutrientscanbewashedawayfromthebodywater 6 byemissaries,atrateaandtherearefurtherlossesduetonutrientsuptakebytheplants,thatneedthemforsurviving. 0 Theuptakerateisdenotedbyb. Theplantsreproducebecausetheyassumenutrients,thisgrazingbeingexpressedby . 1 theproductNP,butonlyafraction0 < e < 1oftheassumednutrientscontributetoincreasetheplantbiomass. The 0 plantshavea naturalmortalityrate m, andexperienceintraspecificcompetitionatrate c. Thedetritusoriginatesby 7 deadplants. Aproportion0 < g < 1oftheplantsthatdiesinktothebottomofthewaterbodyandbecomedetritus. 1 : Inpartthelattermaybewashedawaybyoutgoingstreams,atratew;inaddition,usingdissolvedoxygen,detritusat v rateumaybedecomposedbymicroorganismsintoitsbasicelements,whichcontributetotheformationofnutrients. i X Finally,dissolvedoxygenmaybecontainedintothestreamsflowingintothelake,weassumethistooccuratrateh, r butit is also producedby the aquatic plantsthroughphotosynthesis, at rate s. Because most of aquaticplantsfloat a onthesurfaceofalake, largeamountoftheproducedoxygenmaygointotheair, justifyingthefactthatthe rate s maybesmall. Inaddition,therecouldbeashadingeffect,forwhichonlytheplantsnearthesurfacegetenoughlight, whiletheonesinthelowerlayersofwaterareinpartshadedfromdirectsunlightbythetopmostones. Inorderfor photosynthesistobeviable,plantsneedclorophyllandconsumenutrientsforitsproduction. Thus,inthemodelwe mustmodelthisprocessviathebilinearterm PN. Further,thereisanaturalupperlimitontheiroxygenproduction capacity, that producesa saturation limit, a fact that is expressed by using a Holling type II functionto modelit, k beingthehalfsaturationconstant.Inaddition,dissolvedoxygenisalsowashedawayatratev. Otherlossesaredueto thedecompositionofdetritusintoitsbasiccomponents,aprocessthatwesaidrequiresmicroorganismsthatconsume 1 oxygen,atratez. Withtheseassumptions,themodeltakesthefollowingform: dN dP =q+ fuOD−aN−bNP, =ebNP−mP−cP2, (1) dt dt dD dO sPN =g(mP+cP2)−wD−uOD, =h−vO+ −zOD dt dt k+PN 3. Boundednessofthesystem’strajectories Wenowevaluatetheregionofattractionforallsolutionsinitiatinginthepositiveoctant. LetT = N+P+D+O. Addingalltheequilibriumequations,weobtain d PN T =q+h−aN−m(1−g)P−c(1−g)P2−b(1−e)NP−u(1− f)OD−zOD−wD−vO+s . dt k+PN Droppingsomeofthenegativeterms,takingintoaccountthat0≤e, f,g≤1andthatthelastfractionissmallerthan oneanddefiningβ =min{a,m(1−g),w,v}andH =q+h+s,theabovedifferentialequationsbecomesthefollowing m differentialinequality d T ≤q+h+s−aN−m(1−g)P−wD−vO≤ H−β T. m dt ThesolutionoftheassociateddifferentialequationisT = Hβ−1(1−exp(−β t)+T(0)exp(−β t). Thusweobtainthe m m m finalestimatethatdefinestheregionofattractionforallsolutionsinitiatingintheinteriorofthepositiveorthant: T(t)≤max{Hβ−1,T(0)}. m 4. System’sequilibria Themodel(1)hasonlytwoequilibria:theenvironmentcontainingonlyoxygenandnutrients,E =(qa−1,0,0,hv−1), 3 andthesysteminwhichallthecomponentsareatnonzerolevelE∗ =(N∗,P∗,D∗,O∗). Inadditiontothegeneralcase, thereare severalparticularcases, thatarisewhenoneorbothofthe exogenousinputsare blocked: E = (0,0,0,0) 0 arisingwithq=0,h=0,E =(qa−1,0,0,0)wheneverh=0,E =(0,0,0,hv−1)obtainedforq=0. 1 2 Feasibilityofeach E, i = 0,...,3isobvious. Toanalyzecoexistence,fromthesecondequilibriumequationof i (1) N∗ =(m+cP)(eb)−1≥0, NP=P(m+cP)(eb)−1 (2) areobtained.Nowsubstitutingintothefirstequilibriumequation,wefind OD=Ψ(P)(befu)−1, Ψ(P)≡bcP2+(ac+bm)P+am−beq. (3) ThethirdequilibriumequationnowgivesD∗ =[g(mP+cP2)−uOD]w−1. Using(3),wehaveexplicitly: 1 D∗ = Φ(P), Φ(P)≡AP2+BP+C, A=bc(efg−1)<0, B=befgm−(ac+bm)=bm(efg−1)−ac<0. (4) befw C =beq−am. Now,thesignofODonlydependsonΨ(P). From(3),whenC > 0,(4),toguaranteeΨ(P) ≥ 0,if P isitspositive 2 root, we need P ≥ P . WhenC < 0 instead, Ψ(P) ≥ 0 forevery P ≥ 0. To impose D ≥ 0, we study the concave 2 parabolaΦ(P),withroots P ≤ P . ForC > 0, wehavethatΦ(P) ≥ 0for0 ≤ P ≤ P , goingdownfrom(0,C)to 3 4 4 (P ,0).From(3),(4): 4 OD wΨ(P) O (P)= = . (5) 1 D u Φ(P) 2 Toguaranteeitsnonnegativity,wecombinetheaboveresultsforΨ(P)andΦ(P),togetthefollowingconditionand thecorrespondingnonnegativityinterval: (a)C >0, B<0, P ≤P≤ P . (6) 2 4 Letχ(P)=behk+m(h+s)P+c(h+s)P2>0,Π(P)=bek+mP+cP2>0,Γ(P)=Q+MP+LP2,Q=befvw+zC, M = zB < 0, L = zA < 0andthediscriminantofΓbe∆Γ = M2−4LQ > M2. EliminatingNPfrom(2),thefourth equilibriumequationinsteadgives: O2(P)=hh+sPN∗(k+PN∗)−1i(v+zD∗)−1 =befwχ(P)[Π(P)Γ(P)]−1, (7) TofindsufficientconditionsfortheintersectionofO andO forP≥0,westudyeachfunctionseparately. 1 2 O is positive if conditions (6) hold. In the case (a) O increases from a zero at P = P toward the vertical 1 1 2 asymptoteatP= P . 4 The sign of O is instead determined just by the one of Γ. In principle, there is the following possible case 2 guaranteeingthatthe concaveparabolaΓ is positive. (1) Q > 0, M < 0 which impliesa decreasing branchof this parabolaliesinthefirstquadrantbetweentheverticalaxisanditspositiverootγ+,joining(0,Q)with(γ+,0). Correspondingly,thereisapossiblecaseforO beingnonnegativewhenP≥0:itraisesupfromapositiveheight 2 attheorigintotheverticalasymptoteatP=γ+. We now combine O and O to find an intersection in the first quadrant. (1) is compatible with (a). Sufficient 1 2 conditionsfor(P∗,O∗)tolieinthefirstquadrantis (1)−(a): Q>0, B<0, C >0, P2 < P∗ <P4 <γ+. (8) 5. Stability LetJbetheJacobianof(1)andletJ denotethematrixMevaluatedateachequilibriumE: i i −a−bP −bN fuO fuD   J = (ke+sb0PkPPN)2 ebgNm(−k++smkPN2Ng−)2c2PcP −u−O0zO−w −v−−u0DzD  (9) For J the eigenvalues are all negative, namely −a, −m, −w, −v. Therefore, E is stable. J has the following 0 0 1 eigenvalues −a, (beq − am)a−1, −w, −v. From the condition (8), C = beq − am > 0 we can obtain that if the equilibriumE∗exists,E becomesasaddlepoint. J hasfournegativeeigenvalues−a,−m,−huv−1−w,−vsothatE 1 2 2 isalsoalwayslocallyasymptoticallystable.TheeigenvaluesattheequilibriumE are−a,(beq−am)a−1,−huv−1−w, 3 −vandagainthisequilibriumisasaddlepointwhenE∗ exists. ThestabilityofthecoexistencepointE∗ isexamined numerically. 6. Numericalsimulations TheequilibriumE∗ isfeasibleandstablefortheparametervalues: q= 0.5, f = 0.5,u= 0.5,a= 0.05,b= 0.41, e = 0.9, m = .095, c = 0.08, g = 0.9, w = 0.013, h = 0.3, v = 0.08, s = 0.02, k = 0.02, z = 0.025, obtaining N∗ =0.7801,P∗ =2.4107,D∗ =0.3436,O∗ =3.6098. Letting the input rate of nutrients vary, q, from Figure ?? we observe that with increasing q, the densities of nutrients,aquaticplantsanddetritusincrease.Infact,forq=0theirdensitiesbecomezerowithinashortperiod.The oppositeresultsoccurinstead fordissolvedoxygen,the concentrationof whichincreaseswith decreasingq. When q=0,dissolvedoxygenattainsitsmaximumconcentration. Figure??depictsinsteadtheeffectofrateofconversionofdetritusintonutrientsonthecoexistenceequilibrium E∗. AsimilarsituationasforFigure??arises. Withincreasing f,thedensitiesofnutrients,aquaticplantsanddetritus increase,whiletheconcentrationofdissolvedoxygendecreases. 3 7. Discussion The ecosystem could disappear, as the equilibrium E is locally asymptotically stable, not surprisingly. In fact 0 a pond lacking sufficient inputs becomes dry, and all its componentsdisappear. The no-life equilibrium E is also 2 always stably achievable. E and E can be rendered stable if C < 0. Rewritten as bem−1 < aq−1 it means that 1 3 thereducedplantsnatality,i.e. theratioofnewplantsovertheirmortality,shouldbeboundedabovebythereduced nutrientsdepletionrate, i.e. the ratiooftheirwashoutrate tothe netinputrate. Hence, morenutrientsare supplied intothe systemthan theyare usedforplantreproduction. Insuchcase the nutrients-onlyequilibriumandthepoint withonlynutrientsanddissolvedoxygenarestable. Density of Nutrients0001111.......124682468 qqq===000..58 Density of Aquatic plants123...123555 qqq===000..58 Density of Detritus0000000.......2345678 qqq===000..58 Concentration of DO33333.....456789 qqq===000..58 0.2 0.5 0.1 3.4 00 50 Time/day 100 150 0 0 50 Time/day 100 150 00 50 Time/day 100 150 3.30 50 Time/day 100 150 Figure1: Variationsofnutrients,aquaticplants,detritusanddissolvedoxygenasfunctionsofq. Areductionofthe exogeousinputofnutientslowerstheequilibriumvaluesofalltheecosystemcomponentsbutfordissolvedoxygen. 2 333...246 00..78 fff===000..58 3.49 fff===000..58 Density of Nutrients1.15 fff===000..58 Density of Aquatic plants2222....32468 fff===000..58 Density of Detritus0000....3456 Concentration of DO3333....5678 ff==00.5 2 1.8 0.2 3.4 f=0.8 0.5 0 50 100 150Time/da2y00 250 300 350 1. 60 50 Time/day 100 150 0.1 0 50 100 150Time/da2y00 250 300 350 3.30 50 Time/day 100 150 Figure2:Variationsofnutrients,aquaticplants,detritusanddissolvedoxygenasfunctionsof f. Reducingtheconver- sionrateofdetritusintonutrientsenhancesdissolvedoxygenbuthindersnutrients,aquaticplantsandunexpectedly alsodetritusitself. 1.28 22..389 hhh===000...135 11..112 67 Density of Nutrients111...1246 hhh===000...135 Density of Aquatic Plants22222.....34567 Density of Detritus00000.....56789 hhh===000...135 Concebtration of DO345 hhh===000...135 0.8 22..12 00..34 2 0 50 100 150Time/da2y00 250 300 350 20 50 Time/day 100 150 0.2 0 50 Time/day 100 150 1 0 50 100 150Time/da2y00 250 300 350 Figure3: Variationsofnutrients,aquaticplants,detritusanddissolvedoxygenasfunctionsofh. Whilenutrientsand aquaticplantsarescarcelyaffectedbychangesinthisparameter,forsmallervaluesofhwefindasharpincreasein detritusand a correspondingsensible decrease in the dissolvedoxygento verylow concentrations. To maintain an inflowofdissolvedoxygenappearsthustobenecessarytokeepthewaterbodyhealthy. Acknowledgements Thisworkhasbeenpartiallysupportedbytheprojects“Metodinumericiinteoriadellepopolazioni”and“Metodi numericinellescienzeapplicate”oftheDipartimentodiMatematica“GiuseppePeano”oftheUniversita`diTorino. 4 References [1] A.K.Misra,NonlinearAnalysis:ModellingandControl,12(4),511–524(2007). [2] A.K.Misra,NonlinearAnalysis:ModellingandControl,15(2),185–198(2010). [3] A.K.Misra,AppliedMathematicsandComputation,217,8367–8376(2011). 5

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.