Guild Predation

Home , Bacterivore, Grazing (behaviour), Guild (ecology), Intraguild predation

yeo ril:Articles article: of Type Thisarticle isprotected bycopyright. Allrights reserved doi: 10.1002/ecy.3012 lead todifferences between thisversionandthe been t This article has beenaccepted forpublication and undergonefullpeer reviewbuthasnot hrhsre,AnAbr ihgn S.Eal [email protected] Email: USA. Michigan, Arbor, Ann street, Church 2 1

Accepted USA Michigan, Arbor, Ann street, Church 440 Michigan, of Article University Sustainability, oprtv nttt o ra ae eerh(IL) colfrEvrnetand Environment for School (CIGLR), Research Lakes Great for Institute Cooperative colfrEvrnetadSsanblt,Uiest fMcia,40Cuc street, Church 440 Michigan, of University Sustainability, and Environment for School Intra-guild † ubro iue:4 figures: of Number 40 references: of Number text: main in Words 287 Abstract: in Words hrough thehrough copyediting, typesetting, paginationand proofreadingprocess,whichmay orsodneato:Sho o niomn n utiaiiy nvriyo ihgn 440 Michigan, of University Sustainability, and Environment for School author: Correspondence rydniydpnigo h tegho IGP of strength the on depending density prey egHu Chang Feng-Hsun rdto IP a nraeo decrease or increase can (IGP) predation ∼ 4916 n ro,Mcia,USA Michigan, Arbor, Ann 1, † n rde .Cardinale J. Bradley and Version ofRecord 1 G increase IGP . Please cite this articleas rdces rydensity prey decrease or ,2 1, 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 ouaindnmc;peain protozoa predation; dynamics; population complex more potentially is thought. density previously resource than on its IGP on of depending density impacts resource the Consequently, basal increase our strength. or the Consequently, decrease freed indeed density. can then bacterial IGP which that increased shows prey, work thus IG and the predation by from satiated resource became bacterial predator IG strength IGP the As further, density. increased resource increased turn bacterial in the which decreased and population, density prey consumer IGpredator IG summed the the the of proportion increased small first a IGP consuming by that wewere density suggests model, predictions IGP model with apublished results of empirical ca. 30%)asthe version (by modified explain a to able Using then increased and increased. 25%) IGP ca. of found We strength strength. (by IGP IG decreased for first proxy the a density of as bacterial predator percentage that IG the an bacteria to manipulated available three was experimentally that and We population IGP prey resources. in engaged basal the species increase. We as protozoa ofIGP two species with system asthestrength well-studied a the basal then increase used that to and prediction leads this decrease IGP test first intermediate We would IGP. to density strong resource weak or that weak to so compared density, density prey resource density IG lowest predator the IG reduces the it increases IGP than intermediate more ofIGP to weak asthestrength because occurs then increase This and work increases. theoretical decrease, recent first some should density However, resource density. that resource predicts increase to believed widely is that In Abstract This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article osmrcmuiis nr-ul rdto IP sacmol bevdinteraction observed commonly a is (IGP) predation intra-guild communities, consumer ewrs lpaim;Cliim optto;itagidpeain microcosms; predation; intra-guild competition; Colpidium; Blepharisma; Keywords: ∼ 0 ftevraini rtzaadbceildniy gemn fthe of Agreement density. bacterial and protozoa in variation the of 70% 2 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 because consumer assemblages will become less efficient in consuming their basal resources. their inconsuming efficient less become will assemblages consumer because determine interactions their and consumption. species resource consumer basal make how communities understanding consumer communities in for natural IGP of important of majority uniqueness IGP the and prevalence in The IGP in 2004). al., al., engage et et taxa Thompson (Dunne engage of 2004, more taxa Marquet, or consumer 50% and of that (Arim half and systems than 2007), aquatic more and that terrestrial reported across been IGP has in It 1989, al., 1992). et (Polis Holt, resources (the basal and one shared Polis for another competes on also feeds it which predator) with IG prey) IG intra-guild, intra-guild, (the IGP species 2018). consumer al., et one (Barnes when important and occurs widespread most the of one is (IGP) predation guild we areto if resources. prey partitioning of of resource consumption the controls effects (Sihetal., what the understand with along better consumer community considered ef- bya be how to influence need also and andconsumed interactions 1998), complex arecaptured more These resources prey 1992). Polis ficiently Holt (IGP; and predation Polis intra-guild 1989, and al. 1999) et 1998, Denno, and 1998). Thesecom- (Losey predator- al., 1998), facilitation etal., et predator (Sih (Sih modifications interaction capture predator-prey include interactions onresource 2008). However, plex actthe counter partitioning or of resource and Snyder, enhance caneither effects positive Finke that interaction 2001, inter-specific of Harvilicz, types and several resource are turn,reduces (Duffy there in level and, lower a resources to incapturing consumer density a efficient use, more resource be their to partitioning tends species community consumer When 1993). Schluter, MacArthur, that 1961, 1958, 1957, mechanism (Hutchinson, primary resources the for be competition minimize to to proposed consumers allows been has time or space in partitioning Resource In This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved.

Accepted Article troduction hnIPocr,ms hoeia tde rdc htrsuc est ilincrease will density resource that predict studies theoretical most occurs, IGP When intra- consumers, characterize that interactions complex of types various the Among 3 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 54 53 52 nraebslrsuc est Pese ta. 05.Rcnl,Cage l 09(under 2019 model,they al. simpleconsumer-resource a et Using Chang Recently, explanation. potential 2005). another offered al., review) et (Preisser density and thus resource IGprey basal on some increase effects types, trait-mediated ecosystem induce might to behavior decreases feeding addition species’ In IGP predator IG of 2007). al., occurrence et predator, (Vance-Chalcraft and density prey resource in basal IG IGP of than dynamics important more population is the density. competition governing resource exploitative basal where on ecosystems IGP factors in of Several impacts example, For heterogeneous preydensity? the explain IGP decreases to proposed been that have find others while density, prey increases occurs. IGP when resource basal of density 48% decreased yet, showed 2007); also al., reviewed et studies (Vance-Chalcraft that of IGP found with meta-analysis increases subsequent generally A density resource 2006). basal density Harmon, resource and basal (Rosenheim higher, thanhalf(17 showed occurs more others IGP the while when because lower, showed studies the density of basal resource 29) on of out showed effects Harmon non-significant and had Rosenheim meta-analysis, IGP a that in example, For density. resource basal on prey also to predator or prey IG 2007). allow (Rudolf, (iv) themselves that on or IG 2007), to al., supplement et trophic (Daugherty IG additional predator (iii) than or 2002), prey other (Hart, species resource additional basal (i) and (ii) include predator 2003), that IG prey, al., models et complex (Kuijper more responses in functional same increase nonlinear and always the will competition to both qualitatively IGP due remains that prediction density The resource density 1997). a Polis, lower and had (Holt pressure IGprey, which consumer, consumption the efficient more because the resources consumer be basal the should inconsuming occurs, IGP efficient when less that predicted be model would community IGP classic The 1997. Polis and Holt The This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article aoiyo hs hoeia tde r ae ntecascIPmdldvlpdby developed model IGP classic the on based are studies theoretical these of majority h si hteprclsuishv rvnhtrgnos ihsm idn htIGP that finding some with heterogeneous, proven have studies empirical that it is Why IGP of impacts positive expected the out born always not have studies empirical But 4 103 102 101 100 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 h xeiett rdcin fCage l’ G oe n elzdta hl h two the while agreement poor that the realized suspecting Third, and model poor. was IGP agreement al.’s quantitative agreed, et qualitatively from Chang data of fit predictions we Subsequently, to to resource. experiment system basal the aforementioned the of the density in the IGP impacts this of how strength determine the manipulated we which in experiment strength the vary to order in predator IG the IGP. to of of accessible percentage were the that manipulated population experimentally prey IG we the system, this Using consortium bacterial resources). common basal a a for (the consumed competed that they which predator) The with IG prey) IG (the strong. (the protozoa bacterivore to omnivorous strict moderate an weakto of from from composed grows was IGP IGP grows system is as study as density increasing resource then decreasing basal and IGP-first that strengths, of intermediate strength prediction the the of test function to concave-up experiment a an run to 1999) (Morin, density. resource has basal knowledge affects our IGP to of study hasyetto strength empirical the no how prediction Indeed, investigated this explicitly system. However, biological studies. real negative any times with other tested inempirical be and density positive, resource has on basal sometimes IGP why effects for might assemblages explanation consumer plausible among (under a IGP al. of be the et strength has Chang the thus in results, variation these and that Given suggested pressure, intermediate. review) predation to weak In highest is the increases. IGP to density when density, subjected consumer lowest summed is the resource that basal increases such predator the prey, IG turn, IG the of strength, decrease in the intermediate to than weak more ofthe is function density. IGP up resource When concave is IGP. a on basal of strength density effect basal resource negative that or predicted positive model a their Specifically, has IGP whether control can time, sho This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article e o h tegho G,ie ubro Gpe osmdb nI rdtrper predator IG an by consumed prey IG of number i.e. IGP, of strength the how wed u ae sognzdacrigt h euneo u eerh is,w a an ran we First, research: our of sequence the to according organized is paper Our bacteria consuming protozoa of system study well-developed a used we study, this In 5 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 oo,Cliimsrau I ryadasrc atrvr)adBehrsaamericanum Blepharisma pro- and two bacterivore) and strict prey, a resource and basal prey (IG the striatum as Colpidium served tozoa, that subtilis) Bacillus and cereus, Bacillus ofIGP as the strenght increasing then decreasing, first concave-up - a IGP was increasd. state of steady strenght at the density Polis, of resource) and (Holt (basal function rate bacteria attack if i.e. tested time, modelthatdetermines then of IGP We unit per 1997). classic predator ofthe a by consumed prey components the of of number one the of is Theprobability aprey IGprey. an finding find predator would a predator IG is an consumption that probability for available the altering prey to IG akin of (hereafter, proportion predator the IG Manipulating the prey). by IG consumption of for availability available were that population proportion prey the IG altering the protozoa-bacteria by of accomplished this was which used IGP, We of strength the and 1999). manipulate prey Morin, to IG system 2014, how Morin, examine to and to and (Banerji previously 1993), coexist Morin, used and predators been (Lawler webs has food that of system stability the protozoa-bacteria study a used we experiment, the For Methods 1 Experiment Method concave-up a be IGP. tool to of a density strength as resource the caused model of likely the function that use mechanism then biological to the us deduce allowed between to which agreement predictions, quantitative model and and qualitative we modifiedthe results both empirical in response, resulted IIfunctional This Type a accordingly. model fact, follow after in Fourth, do, consumers. the consumers of the response confirming functional the characterize to experiment second a w This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved.

Accepted Article performed we model, the in used response functional I Type simplistic overly the to due as h oa raim nteeprmn eetrebcei pce Srai marcescens, (Serratia species bacteria three were experiment the in organisms focal The 6 153 152 151 150 149 148 147 146 145 144 143 142 141 140 139 138 137 136 135 134 133 132 131 130 129 128 oiino h ls ote ecetd2% 0,6% 0 G teghtetet,each treatments, strength times. IGP 5 80% replicated 60%, was 40%, which 20%, of created 20- we the bottle, installing glass By the IGP. of of strength the position the of thus, ratio and prey The predator, IG of IG available. availability the the not to represented was bottle prey experimental entire to IG the available the to was relative where space prey space feeding IG refuge the a which and in predator, space IG feeding the a spaces, two into 20- the bottle the of experimental in location except The unit treatment. experimental entire IGP the 0% in distributed homogeneously be should bacteria 20- the that by confirm constrained to was 20- microscope predator the used through had pass we could predator, 20- prey IG the IG the the that not but reassert prey To IG the to bottles. permeable experimental other the the For in predator. installed IG was the to available 20- was population treatments, prey 4 IG the of 100% 10- that the such replacing by created 250- was a treatment strength with IGP mesh 100% The predator. IG To the mesh to Cloth-Nitex 1). Bolt (Figure installed the predator of IG size 10- mesh the was the to treatment, available strength made IGP 0% being the population create prey IG the of 100% between. mesh in aBoltCloth-Nitex installed with next) described glued together size, then varying were (of which and off, cut bottoms their cultured had were species exhibit focal to The known not 1999). in are (Morin, but cannibalism IGP, like in relationships engage feeding to known other are species protozoa two The 1999). (IG This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved.

Accepted Article∼ rdtradaonvr)ta evda h osmr BnriadMrn 04 Morin, 2014, Morin, and (Banerji consumers the as served that omnivore) a and predator 0m xeietlbtlsta eemd rmto20m opkgasbtlsthat bottles glass Qorpak mL 240 two from made were that bottles experimental 300mL h xeieticue ramnsrpeetn % 0,4% 0,8% and 80%, 60%, 40%, 20%, 0%, representing treatments 6 included experiment The µ ,wihwsntprebet ohcnuess htn Gpe eeavailable were prey IG no that so consumers both to permeable not was which m, µ µ ehta a emal obt h Gpe n h Gpredator, IG the and prey IG the both to permeable was that mesh m ehta a emal oteI rybtntteI predator IG the not but prey IG the to permeable was that mesh m µ eh hrfr,teI ryadteee smaller even the and prey IG the Therefore, mesh. m µ ehwtotdfiut u h oeetofIG the movement but difficulty without mesh m µ 7 ehwsmnpltdt iieteentire the divide to manipulated was mesh m µ µ ehwsindeed was mesh m ehi different in mesh m µ m 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 154 xeietlssesrahdsed tt ogl 9 o48husatriouainof inoculation after hours S1). 468 state. Figure to the steady S1: 298 as (Appendix roughly protozoa state defined time was steady The density reached window. protozoa systems time in Experimental previous change the least of in the density that showed mean by the that window divided window and time 468, present hour the to in 34 window protozoa hour time from both five-point time the a moved at gradually point five consecutive time then across one We protozoa forward treatment. of both IGP each for mean density assess points the To time density. calculated population first protozoan we to state, respect steady with state steady reached units mental microscope. 500 another and density prey) (IG striatum Colpidium count to used 100 the density, was protozoan to media count back subsampled To once supplied volumn. was species total media the protozoa sterile maintain to both fresh bottle mL of experimental 5 subsampled density subsampling, was the each total After count in day. to media other bottle of every experimental mL 5 of density, side protozoa both track To from weeks. 4 for day other Celsius. 20-degree to set was temperature where shakers chamber the growth and a (rpm), inside minute placed per rounds were Orbital 60 Benchtop under 2000 suspended organisms MaxQ the Scientific keep Thermo to Shakers the on placed protozoan were the protozoa and cultured unit), the with experimental to the of added side then each were on 200mL species (one seeds were inoculated. wheat species 2 bacteria media, thethree the of which, sterile after L flask, 1 1400-mL in a in Carolina) water North DI Burlington, Supply, Biological (Carolina pellets” ”protozoan This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. µ

Accepted dissecting 4X under density predator) (IG americanum Blepharisma count to used was L Article Media enx esrdbcei este tsed tt ots h yohssta bacterial that hypothesis the test to state steady at densities bacteria measured next We experi- the when determine to densities protozoan for data monitoring the used We every species protozoa two the of density the monitored we experiment, the During o utrn rtzaseisi h ote a rae ydsovn .7mg 0.07 dissolving by created was bottles the in species protozoa culturing for ∼ 0 Leprmna ote h xeietlbottles experimental The bottle. experimental mL 300 8 µ fthe of L 205 204 203 202 201 200 199 198 197 196 195 194 193 192 191 190 189 188 187 186 185 184 183 182 181 180 010 ramns iue2.Qaiaiey hs xeietldt ac h theoretical the match data experimental 0-20%and these between Qualitatively, (p 2). treatment density Figure IGP test comparing treatments; 0% 80-100% the Difference Significant to Honest relative Tukey’s 36% for by 0.02 increased = density increased bacterial further 100%, IGP and of strength 80% the to be- as However, density test comparing treatments). 40-60% and Difference by 0-20% Significant decreased the Honest resources) tween Tukey’s (basal for bacteria 0.05 of strength = density the (p the As 25% 60%, roughly 2). ca. Figure to IG of 0% line the from (solid increased of IGP predator 37% of IG of the strength to IGP available an being at population occurred prey function quadratic the 1; standard error of = minimum internal term (constant quadratic apositive had 0.14; data = ifthebacteria the func- examine to quadratic to The fitted data IGP. best the of that strength to tion the versus relationship function concave-up a aquadratic exhibited density fitted first we state, steady At Results Team,2018). (R 1 Core Experiment 3.5.2 R by done were theincrease exercises with fitting The then increase and strength. IGP decrease of thanzero,we first greater would density bacteria term termwas significantly the thequadratic that with concluded strength the constant and IGP zero associated the coefficient than within the less significantly data tostatistically If located was were manipulation. the density our to of bacteria (0-100%) of range values function extreme aquadratic the if fitted examine and strength IGP versus density Gibco with diluted after value re-run recommended be the than higher was sample Attune a the in by density cell the the if addition, for In samples clogging. prepare To 20 through density. passed protozoa were samples to all cytometer, respect with state steady reached system Attune a used we densit This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article is erae n hnicesswt G tegh oqatf bacterialdensities, quantify To IGPstrength. with then increases and decreases first y p < 0.01 ® cutcFcsn yoee aul(> manuel Cytometer Focusing Acoustic udai em=14,sadr ro 0.6, = error standard 1.42, = term quadratic ; ® cutcFcsn yoee ocutbcei fec elct fe the after replicate each of bacteria count to Cytometer Focusing Acoustic TM B ufr.Fnly epotdteosre bacteria theobserved weplotted Finally, buffers. PBS 9 µ eht eoelreprilsadt avoid to and particles large remove to mesh m 10 6 el e L,tesml would sample the mL), per cells p = 0.03 ; R 2 = 0.54 .The ). 230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 hsadtoa xeietwsrni 0m daee)x1 m(egt Fisherbrand (height) mm 15 x (diameter) mm 60 in run predator. was IG experiment by prey additional to IG This experiment of consumption additional the functional an describing curve performed of response we functional type the species, quantify the protozoa determine two II the To Type by by exhibited 1975). approximated consumption response Stewart, better the and be suggested (Laybourn may have Blepharisma response authors and functional some Colpidium I, both Type by consumption simple bacteria all the of model to to contrast response functional In I Type terms. that simplistic was overly review) an (under 2019 used authors al. et the 1 Chang Experiment of in predictions measured model match densities quantitatively bacterial not why did reason likely most the that suspected We Methods 2 Experiment predictions. revisions that theoretical model and and data work empirical of match experimental of better additional Because a topursue achieve 2). would Figure decided in we line model dotted fit, between 65.05; poor = difference the square the of (sum increase data further empirical values and parameterizing literature predictions Unfortunately, with 1). model Toimprovethe (Table al.’s toparame- data. literature et was the from Chang values of attempt with mean first al’s our et grand Chang results, the terized empirical than and data predictions model the empirical between to match review) (under 2019 fit al. poorer et squares Chang a of from sums prediction was total model the the means than which greater (0.365), was data 0.697) the predic- = of model square between of (sum difference our data The which empirical from line). and review), tions dashed (under long 2019 2 no- al. (Figure derived et dots) were Chang solid predictions of 2 output (Figure model data the empirical from diverged the tably prediction, priori a our matched qualitatively density. resource basal increases then and prediction This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article lhuhteeprcldt nbcei est tsed tt rmeprmn 1 experiment from state steady at density bacteria on data empirical the Although rmCage l 09(ne eiw httesrnt fIPfrtdecreases, first IGP of strength the that review) (under 2019 al. et Chang from 10 TM 255 254 253 252 251 250 249 248 247 246 245 244 243 242 241 240 239 238 237 236 235 234 233 232 231 olwn oreutost ecietepplto yaiso w aa eore swell as resources the basal predator. used two and we of prey review), dynamics IG (under population as 2019 the describe al. to et equations Chang four as following structure model general same response the functional Using II Type with model IGP revised A published values with or time) handling and literature. rate the with attack in model IGP revised (the (under the 2 2019 parameterized Experiment then al. of and et results response, Chang functional in II terms Type consumption a all with modify review) results, to these was Given step respectively. next 0.36, the and decided 0.39 we be to estimated were time handling and in rate line (dashed ( response response functional functional initial linear saturating the II against Type day a per 3a.; that predator Figure found IG we per density, consumed prey prey IG IG of number the plotting By Results - 2 Experiment per predator IG per consumed prey IG of number intreatments the day. were changes predator density IG the between without treatments differences and in The with prey Celsius IG predator. of 20-degree IG changes same without density the the and recorded with in we rpm hours, 24 60 Scientific After Thermo at chamber. the rotating growth on Shakers also Orbital placed protozoa-free Benchtop then of 2000 were 15.11, mL 9.89, MaxQ dishes 0 4.11, Petri was 2, beginning The 4, the individual/mL. in 6, levels 29.78 8, 5 the 20, with of prey) prey IG (IG of Colpidium density 2,4, average with mixing The media by medium. of werecreated mL prey 10 density IG IGprey and of of 8, levels 5 6, levels same five the The and predator. individual, IG predator without IG density one with density prey IG of levels 5 P This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved.

Accepted Article representing treatments 10 of each for units replicate 3 up set We Lid. Clear with Dishes etri R 2 = 0.91 a etrepaaino h nr-ul rdto hnTp I Type than predation intra-guild the of explanation better a was ) p < 0.01 .Fo h yeI auaigfnto,teIPattack IGP the function, saturating II Type the From ). 11 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 olwn loaTp Ifntoa epnewt h G takrt ( predation, rate intra-guild attack IGP the the with i.e. response prey, functional IG II the Type consumed a also also following predator ( IG efficiency the assimilation terms, the consumption addition, multiplied the In response, by functional determined II was Type predator a IG followed and now which prey IG of rate Growth 4. and on consuming When specialists complete resources. a was both on equally consuming When generalists predator. complete (under and 2019 prey IG al. among et partitioning resource Chang Following respectively. parameter predator, the IG review), or prey IG indicating 2 or ( predator IG rates growth intrinsic capacities with carrying logistically grew resources basal two The ( This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. dZ dZ dR d R d d d d R t t t t 1 2 1 Accepted Article2 1 R and = = = = 1 nacrac ihCage.a.21 udrrve) h yaiso atraspecies bacteria of dynamics the review), (under 2019 al. et. Chang with accordance In h yaiso Gpe ( prey IG of dynamics The . [ [ r r 2 1 1 1 R R R + + e 2 1 e 1 2 2 ( ( h h c 1 1 c ,I ry( prey IG ), 2 1 1 2 − − c c sR ( 2 1 1 Z ( s K K R R 1 − 1 R 2 2 1 2 1 + olwn yeIfntoa epnewt takrt ( rate attack with response functional I Type a following ) 1 − s ) ) + ) e − − s R 1 ) K c h 1 R [ [ 1 1 + 1 ( 1 1 1 c 1 1 Z + + + and e ( − s 2 1 1 h h h agn rm05t ,wsdsge omnplt h ereof degree the manipulate to designed was 1, to 0.5 from ranging , c n rdtr( predator and ) − s 2 2 1 1 ) sR c c c R 2 1 1 s K ) s s s c 2 2 R R R R 2 1 + ( ohbslrsucswr osmdb Gpe ( prey IG by consumed were resources basal Both . 2 2 1 1 1 c ] + + + 1 e − Z 3 s R c h 1 h h 3 s 3 1 1 1 − Z ) α c c c R Z 3 1 1 1 [ α ( ( 2 1 n Gpeao ( predator IG and ) 1 1 1 R Z + − − 1 1 n h rdtr( predator the and , ] Z h s s Z ) ) 2 2 2 R R c eedsrbdb qain1t respectively. 4 to 1 equation by described were ) 2 − 2 2 12 ( ] ] 1 mZ Z Z − 1 1 − − 2 s , ) R [ [ 1 1 c 1 3 + + + α Z h h h 1 2 2 2 c c c Z 2 2 2 s s ( ( 2 1 1 R eedsrbdb qain3 equation by described were ) = Z 2 − − 2 + 0.5 s s a opeeyspecialized completely was ) c ) ) h 2 R R 3 ( ohcnuesbecome consumers both , 1 1 c 1 c 3 2 + + s − α sR Z = h h s 2 1 2 2 ) r c c R ] 1 1 2 2 Z 1 c s s and 2 R R 3 h Gpe ( prey IG the − ,hnln time handling ), 2 2 c + + i mZ ,weei=1 = i where ), r h h 2 3 3 1 swl as well as , c c , 3 3 α α Z Z Z 1 1 1 ] ] and ) Z Z 2 2 Z e , , (4) (3) (2) (1) i 1 ). ) 296 295 294 293 292 291 290 289 288 287 286 285 284 283 282 281 280 279 278 277 276 275 274 273 272 271 n h est fteI rdtra taysaei xeiet1 epciey(iue4). prey (Figure IG respectively the 1, experiment of in density state the steady at as predator well IG as the density, of density bacterial the summed and for variance the of 66% (R Core Team, the 3.5.2 R by and 2017) done was Inc., 2018). data Research, andempirical (Wolfram 11.1 predictions model Mathematica between of difference aid the with generated was this, ( doing efficiency By assimilation ( partitioning of resource S2). combination value Figure parameter S1: best Appendix the found and we S1, (Appendix behaviors of cycle steps limit resources, time two exhibit 2000 of that (last steps) Note average time predictions. simulated long-term 10000 model the the the from of calculated square were of predictions sum model the residual the from the predictions calculate with to overlaid model were II patterns at Type empirical predator These and prey predator. IG IG to ( of available density strength wefirst the IGP data, as against well state as andempirical steady bacteria of predictions density model summed between the difference plotted andempirical the estimate predictions To model between difference data. least the yielded that parameters two of ( two ( toIG predator of predator exception prey ( with IG partitioning 1), resource Table of from of efficiency column 3a) assimilation Figure model the time; II handling parameters: (Type and literature rate attack published IGP the (the and 2 experiment from values with model ( mortality independent density a had predator and prey IG The predator. IG to ( This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. h

Accepted Article 3 ), siiainefcec ( efficiency assimilation ognrt rdcin rmteTp Imdl(qain -) eprmtrzdthe parameterized we 1-4), (equations model II Type the from predictions generate To h eie G oe ihaTp Ifntoa epneepand6% 8,and 68%, 67%, explained response functional II Type a with model IGP revised The e 3 n h ereo eorepriinn ( partitioning resource of degree the and ) α s ob 0 n 6 epciey(iue3) h oe predictions model The 3b). (Figure respectively 96% and 40% be to ) stesrnt fIPta stefcso hssuy ial,both Finally, study. this of focus the is that IGP of strength the is s .W siae h siiainefcec rmI rytoIG prey IG from efficiency assimilation the estimated We ). e 3 ,adteparameter the and ), α ,ie % 0,4% 0,8% 0%o Gprey IG of 100% 80%, 60%, 40%, 20%, 0%, i.e. ), 13 Z 1 α and eciigteaalblt fI prey IG of availability the describing s ysacigfrtecombination the for searching by ) Z 2 m eas h oe per to appears model the because 1 and m 2 ). e 3 n h degree the and ) e 3 n ereof degree and ) 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 oeta ilgclmcaim htmgtudri h atrsosre neprmn 1. experiment in observed patterns the identify underlie helped might model that II mechanisms Type biological The potential work. values, experimental literature own with our parameterized by partly parameterized was partly model and That 4). Figure in lines (solid sponse ( well reasonably al., et (Vance-Chalcraft opposite the of IGP shown have 2007). effects others negative whereas demonstrated density, have resource Thisfinding basal studies resource. on empirical of basal some why explain the density help predation then increases could intra-guild and that review) decrease (under first supports 2019 would finding strength al. This et 4a). Chang Figure and then from and prediction 2 theoretical (Figure decreasing, the increased first IGP - of strength predation the intra-guild as of increasing resource function prey basal concave-up of a density the was state, (bacteria) steady the at that, showed results experimental Our Discussion of response discussion). functional the II in Type further saturating (explained IGP the in from involved resulted steady consumers at 4a) density Figure bacterial of and the increase 2 then from (Figure and predictions decrease state the and that data infer can empirical we between model, II match off the Type improved leveled and then the model and Given II and Type increased 4c). experiment the actually (Figure the density, strength in predator IGP observed IG that the was suggested for data that Finally, experimental 4b) 4b). Figure Figure 0% in in from line dots decrease solid (solid monotonic (the the prey strength predicted IG also IGP the model 100% For II to 4a). Type (Figure the IGP state, of steady strength which at the beyond density with predator) increase IG to to started population density prey bacteria IG of availability (40-60% density bacteria The This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article oe ihaTp Ifntoa epnemr prpitl atrdtetrsodof threshold the captured appropriately more response functional II Type a with model h bevdpplto yaiso atra Gpe n rdtrwr explained were predator and prey IG bacteria, of dynamics population observed The ∼ 0)b nIPmdlta a oiidt aeaTp Ifntoa re- IIfunctional Type a to have modified was that model IGP an by 70%) 14 347 346 345 344 343 342 341 340 339 338 337 336 335 334 333 332 331 330 329 328 327 326 325 324 323 322 esrd ti osbet ne au sn iesetu hoy(er 94 Sheldon 1974, (Kerr, theory spectrum size using value a infer to possible is it measured, all. quantitative at imperfect none than an better that is feel system we experimental model, the the the of dampen of description potentially limitations could clear values from Despite parameter efficiencies cycles. these conversion limit in and/or reduction rate A in growth lies protozoa. mismatch bacterial to likely of bacteria most measures the the that of speculate We accuracy the model. the from from missing those still aspects matched certain are model suggests biology that the of dynamics cycles temporal of in limit values mismatch by a mean was characterized the clearly there was While experiment, model experiment S2). II the Figure Type of the S1: time-frame whereas (Appendix the over S1), state Figure steady densities a S1: population reached (Appendix predator Indeed, and experiment. mimic prey to the IG failed of the developed dynamics of we that temporal model the II of Type aspects the hand, certain other a the 70% ofthevariation was On model data. the empirical approximately in addition, explaining uprelationship, In concave work. to this was theoretical which fit previous resources, reasonable bacterial on rela- the based concave-up of prediction the density original capture the and qualitatively our IGP to of appear strength did the between model tionship the hand, one On sword. resource 4). bacterial Figure highest and 2 thus (Figure and IG equilibrium consumption satiated at bacterial and density density lowest prey the IG in density of resulted prey level IG lowest predator the the and Consequently, bacteria level. consuming lowest stopped by the predator satiated reached IG be the to started that predator such IG prey the IG further, the increased lowest IGP the of to strength suppressed the was When density con- level. bacteria summed that The such prey. higher IG therefore, the was, density of sumer consumption moderate to due increased density predator When This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article Though the assimilation efficiency from IG prey to IG predator has not beenempirically has toIGpredator prey IG from efficiency assimilation the Though isabitofdouble-edged the experiment from to data model II the Type of fit The h G teghicesdfo %t oemdrt ees(..4-0) h IG the 40-60%), (e.g. levels moderate more to 0% from increased strength IGP the 15 373 372 371 370 369 368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 opdu u uhdfeec hudntcag ourconclusions. not change should difference attack such of but value Colpidium the Consequently, predation. ( intra-guild rate with increase then attack of and value decrease the altering between After match the ( S2). change rate (Appendix qualitatively results the to empirical altering appear and However, predictions not experiments model did species. two our parameters the in two resources these among food of differ same value may the parameters on and Gualtieri, two fed these they (Verni although and feeders 2010) filter al., both et were Thurman species 1997, both because decision this made We ( efficiency experi- withadditional or refined verified be to ments. needs seems parameter resources bacterial particular of this partitioning unlikely, cil- complete similar Because relatively 1980). with (Fenchel, feeders resource structures filter exhibit iary certainly both species are two species con- the two items While these food prey. biologically, in IG partitioning, the separation and 96% predator a IG IG was the the there by between suggested sumed partitioning model Our resource for predator. IG obtained and we prey value the of skeptical somewhat are toourfitted similar is which 2009), etal., 41.7% (Andersen (40%). be assimilation estimate the to 2005), Morin, and ispredicted exponents (Long scaling study our those efficiency in Given ratio exponents size scaling 2006). body Beyer, with predator-prey and law the and power (Andersen a respectively as 0.80, and size 0.75 body of with scale should rate search volumetric et This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved.

Accepted Article and rate consumption individual’s an theory, spectrum size the to According 1977). al., u xeietsfesaltecvasta r yia fmcoom experiments microcosms of aretypical that the caveats all suffers experiment Our ( parameters fitted two the to addition In we ofevidence, lines other with isconsistent efficiency assimilation for value the While c c 2 2 n siiainefcec ( efficiency assimilation and ) n siiainefcec ( efficiency assimilation and ) e 2 fBehrsao atrawr e ob h aea hto Colpidium. of that as same the be to set were bacteria on Blepharisma of ) e e 2 2 fBehrsao atracndfe rmta of that from differ can bacteria on Blepharisma of ) esilfudtedniyo aa eore first resources basal of density the found still we ) 16 e 3 and s ,atc ae( rate attack ), c 2 n assimilation and ) 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 384 383 382 381 380 379 378 377 376 375 374 h mat fIPcudb oecmlxta rvosyepce.Teeoe fw are we if Therefore, expected. that previously suggests than study complex more Our be density. could IGP resource of basal basal impacts increases control the always to the turn, consumers to in of counter ability and, the runs resources interferes strength always its IGP on that depending thinking density IGP can conventional resource that basal finding increase Our or density. decrease resource basal impacts predation intra-guide of strength more in experiments or studies observational obvious using next prediction systems. the theoretical natural the so, specific test done a similarly have test we to to that is Now is step one environment. this controlled like highly experiment in ecosystems. microcosm prediction aquatic real a to of results extend point to we the not Nevertheless, is Rather, study done. this in been goal has our research that out more point before actual would resource the theo- third discern between can’t the We having match of deteriorated. the impacts be scenario, would such results In empirical and change predator. predictions might or retical resource prey third IG the of or density predator population third and the the prey if IG partitioning by resource consumed of evenly degree was the resource decrease might theotherhand, On resource third results. the might of resource andempirical addition third model the resource theoretical having of between (96%), match degree the high the influence very not Since be to predator. estimated or already prey was IG partitioning by the if consumed partitioning mainly resource was of degree resource ofour speculations One the third increased only resource work. third theoretical the ourempirical the including to that in was resource agreement third the model the the in having influence resources how might two model speculate were only there can but We species, experiments. empirical model. bacteria the theoretical three match were to model there havedone the we experiments, the andparameterize although In model, the to design from efforts differs best struc- setup our trophic experimental natural our of addition, lack In as well tures. as interactions, species and variability environmental (Briggs This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article oorkolde hsi h is miia td oepiil netgt howthe investigate toexplicitly study empirical first the is this knowledge, our To n oe,20) hs aet nld ml eprladsailsae restricted scale, spatial and temporal small include caveats These 2005). Borer, and 17 410 409 408 407 406 405 404 403 402 401 n-emDsetto elwhp nvriyo Michigan. Rackham of and University Taiwan Fellowship, Education, scholarship Dissertation of One-Term the Ministry by study, overseas sponsored for is sponsorship work com- government Her This of manuscript. manuscript. the the of strengthened aspects various ultimately regarding ments Moeller us Holly with thank discuss to especially We willingness her consultant. the for theoretical on the feedback and on advice Ke valuable Po-Ju their and for manuscript Godwin Casey and consul- Ibáñez and set-up Inés microcosm thank We on help tant. kindly their for group research Morin’s Peter thank We Acknowledgement predation. intra-guild of strength the quantify explicitly to This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article etrudrtn htcnrl h osmto fbslrsucs emyne to need may we resources, basal of consumption the controls what understand better 18 433 432 431 430 429 428 427 426 425 424 423 422 421 420 419 418 417 416 415 414 413 412 411 uf,J .adA .Hriiz 01 pce-pcfcipcso rzn amphipodsin ofgrazing impacts 2001. Species-specific A. M.Harvilicz. and E. J. Duffy, intraguild to supplements Trophic 2007. Briggs. J. C. and Harmon, P. J. P., M. Daugherty, long-term allow not may experiments short-term Why 2005. Borer. T. E. and J. C. Briggs, I. M. Scherber, C. Eisenhauer, N. Lefcheck, S. J. Jochum, M. D., in- A. an Barnes, in competition apparent Trait-mediated 2014. Morin. J. P. and A. Banerji, related interaction widespread A predation: Intraguild 2004. Marquet. A. P. and M. Arim, efficien- individual and Trophic 2009. Lundberg. P. and Beyer, E. J. H., K. Andersen, Andersen, References This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved.

Accepted 116:662–677. Oikos, predation. 15:1111–1117. Applications, Ecological predation. intraguild about predictions between Link The Evo- and Flux: Ecology S0169534717303257?via{%}3Dihub URL Energy in Trends 33:186–197. 2018. Functioning. lution, Brose. Ecosystem U. and and Biodiversity Multitrophic Ruiter, de P. O’Connor, URL j.1600-0706.2013.00937.x 123:567–574. Oikos, system. predator–prey traguild 7:557–564. Letters, Ecology biology. Article species to Biological B: Society Royal 1098/rspb.2008.0951 the of URL Proceedings 276:109–114. Sciences, communities. size-structured of cies URL 168:54–61. Naturalist, American org/10.1086/504849 The Spectrum. Size Marine the in .H n .E ee.20.Aypoi ieDtrie pce Abundance Species Determines Size Asymptotic 2006. Beyer. E. J. and H. K. . . https://www.sciencedirect.com/science/article/pii/ http://rspb.royalsocietypublishing.org/cgi/doi/10. . . 19 https://doi.org/10.1111/ https://doi. 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 er .R 94 hoyo ieDsrbto nEooia omnte.Junlo the of Journal Communities. Ecological in Distribution Size of Theory 1974. R. S. Kerr, 95:137–145. Naturalist, American The Plankton. the of Paradox The 1961. E. G. Hutchinson, Quantitative on Symposia Harbor Spring Cold remarks. Concluding 1957. E. G. Hutchinson, The Predation. Intraguild for Framework Theoretical A 1997. Polis. A. G. and D. R. Holt, lake in cascades trophic and predators, invertebrate predation, Intraguild Diverse 2002. R. by D. Partitioning Hart, Resource Increases Niche 2008. Snyder. E. W. and L. D. Finke, ecological their and rates Feeding protozoa: ciliated in feeding Suspension 1980. T. Fenchel, 2011. Véscovi. García E. and Colombo, I. M. Venanzio, Di G. Campoy, M. E. V., G. of Fedrigo, robustness and structure Network 2004. Martinez. D. N. and Williams, J. R. J., Dunne, This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. URL 22:415–427. Biology, URL 149:745–764. Naturalist, American 218:111–128. Biology, Theoretical of Journal webs. food 321:1488–1490. Science, Exploitation. Communities URL BF02020371 6:13–25. Ecology, Microbial significance. inside vacuoles autophagic-like 6:e24054. in ONE, proliferate PLoS and cells. survive non-phagocytic to able is marcescens Serratia 273:291–302. Series, Progress Ecology Marine webs. food marine //www.int-res.com/abstracts/meps/v223/p201-211/ an Accepted URL 31:1859–1862. f74-241 Canada, of Board Research Fisheries Article egasbdcmuiy aieEooyPors eis 2:0–1.URL 223:201–211. Series, Progress Ecology Marine community. eelgrass-bed http://www.jstor.org/stab . . . https://doi.org/10.1086/286018 20 http://link.springer.com/10.1007/ . https://doi.org/10.1139/ . https: 480 479 478 477 476 475 474 473 472 471 470 469 468 467 466 465 464 463 462 461 460 459 458 457 oi,G . .A yr,adR .Hl.18.Teeooyadeouino intraguild of evolution and ecology The 1989. Holt. D. R. and Myers, A. complex C. of A., G. dynamics Polis, The predation: Intraguild 1992. Holt. D. R. and A. G. Polis, experimental in dynamics population and predation, intraguild Productivity, 1999. coniferous P. Morin, northeastern of warblers some of ecology Population 1958. H. R. MacArthur, central the predation: synergistic facilitating Factors 1999. Denno. F. R. and E. J. Losey, enhanced interactions: predator–predator Positive 1998. Denno. F. R. and E. J. on Losey, composition and community size of organism Effects 2005. Morin. J. P. and the T. in Z. Growth Long, and Consumption on Studies 1975. Stewart. M. J. and in M. E. dynamics J. population Laybourn, and architecture web Food 1993. Morin. J. P. and P. S. Lawler, Kuijp This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. rpi neatos rnsi clg n vlto,71114 URL 7:151–154. Evolution, and sciencedirect.com/science/article/pii/016953479290208S Ecology in Trends interactions. trophic URL 80:752–760. Ecology, webs. food URL 39:599–619. Ecology, forests. 1051-0761(1999)009[0378:FFSPTC]2.0.COhttp://0.0.0.2 URL 9:378–386. Applications, Ecological synchrony. of role 79:2143–2152. Ecology, populations. URL aphid of suppression synergistic and rates predation 8:1271–1282. Letters, Ecology functioning. ecosystem 44:165–174. Ecology, Animal of Journal Stokes. campylum Colpidium Ciliate 141:675–686. naturalist, American The protists. of microcosms laboratory 163:19–32. Modelling, Ecological dynamics. web food and rdto:ptnilcmeiosta a ahohr nulRve fEooyand Ecology of Review Annual other. Accepted each 20:297–330. eat Systematics, that competitors potential predation: Article r .D . .W oi .Zneed n .A .M oimn 03 Omnivory 2003. Kooijman. M. L. A. S. and Zonneveld, C. Kooi, W. B. J., D. L. er, https://doi.org/10.1890/0012-9658(1998)079[2143:PPPIEP]2.0.CO;2 http://www.jstor.org/stable/1931600 http://www.jstor.org/stable/177014 21 https://doi.org/10.1890/ . . http://www. . . . 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 hmsn .M,M ebr,B .Sazmk,adJ .Sui.20.Tohclevels Trophic 2007. Shurin. B. J. and Starzomski, M. B. Hemberg, M. M., R. Thompson, prey. on predators multiple of impacts Emergent 1998. Wooster. D. and Englund, G. A., Sih, of Food Pelagic 1977.Structure M. A.Paranjape. and Jr., Sutcliffe H. W. W., R. Sheldon, efficiency. use habitat and shape, size, sticklebacks: in radiation Adaptive for 1993. D. Consequences Schluter, omnivory: and cannibalism of interaction The 2007. V. W. H. Rudolf, onthe Predation of Intraguild Influence The 2006. Harmon. P. J. between and Relationship A. J. 1982. Rosenheim, Ball. A. and McMeekin, A. T. Olley, J. URL A., D. Computing. Ratkowsky, Statistical for Environment and Language A R: 2018. Team. Core R Preisser, This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. n rpi age:tepeaec fonvr nra odwb.Eooy 88:612–617. Ecology, webs. food real in omnivory of prevalence the tangles: trophic and 13:350–355. Evolution, and Ecology in Trends URL Fisheries 34:2344–2353. the Canada, of of Journal Board Production. Research Fish and Plankton Between Relationship and Chain URL 74:699–709. Ecology, 88:2697–2705. Ecology, dynamics. community 1–20. pages Brodeur Control, Interactions J. URL Biological Dordrecht. In Netherlands, in Springer Guild Reassessment. and Trophic Empirical editors, An Boivin, G. Population: and Prey Shared a of Suppression 149:1–5. Bacteriology, of Journal cultures. bacterial of rate growth and temperature https://www.r-project.org/ URL 86:501–509. Ecology, https://esajournals.onlinelibrary.wiley.com/doi/abs/10.1890/04-0719 interactions. predator-prey in consumption and intimidation AcceptedURL Article https://doi.org/10.1890/05-1454 .L,D .Blik n .F ead 05 crdt et?teefcsof effects the death? to Scared 2005. Benard. F. M. and Bolnick, I. D. L., E. http://www.jstor.org/stable/1940797 . https://doi.org/10.1007/1-4020-4767-3{_}1 22 . https://doi.org/10.1139/f77-314 . . . . 513 512 511 510 509 508 507 506 505 504 ofa eerh n.21.Mteaia eso 12 hmag,Ilni,USA. Illinois, Champaign, 11.2. Version Mathematica, 2017. Inc. Research, Wolfram 504. – 28:487 Micron, protists. ciliated in behaviour Feeding 1997. Gualtieri. P. and F. Verni, 2007. Sih. A. and Osenberg, W. C. Vonesh, R. J. Rosenheim, A. J. D., H. Vance-Chalcraft, Th This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved.

Accepted ArticleURL Ameta- release: prey and preysuppression on 88:2689–2696. Ecology, analysis. predation of intraguild influence The URL 588. – 161:577 Protist, prey. //www.sciencedirect.com/science/article/pii/S1434461010000234 in bacterial behaviours equally-sized, mixed, feeding of selective presence display the pyriformis tetrahymena and striatum colpidium ra,J,J .Pry .J il n .Lyor-ar.21.Tefle-edn ciliates filter-feeding The 2010. Laybourn-Parry. J. and Hill, J. P. Parry, D. J. J., urman, http://www.sciencedirect.com/science/article/pii/S0968432897000280 23 . http: . 515 514 est needn otlt ( mortality independent Density – ( partitioning resource of Degree – ( efficiency assimilation IGP – ( time Handling IGP – ( rate attack IGP – predator) ( (IG Blepharisma to bacteria from efficiency Assimilation – bacteria ( on predator) (IG Blepharisma of time Handling – bacteria ( on predator) (IG Blepharisma of rate Attack – prey) ( (IG Colpidium to bacteria from efficiency Assimilation – bacteria ( on prey) (IG Colpidium of time Handling – bacteria ( on prey) (IG Colpidium of rate Attack – ( capacity carrying Bacteria – ( – r e c e c e c K h h h i T This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. atraprcpt rwhrate growth capita per Bacteria † 3 3 2 2 1 1 3 2 1 1/day) ; i ; 1/ ; %) ; 1/ ; %) ; 1/ ; ; ; ; ind./mL) ; dat This dat day day Accepted Article able ta.21 udrrve)admdlwt yeI ucinlrsos nti study. this in response functional II Type with Chang model in and response review) functional (under I 2019 Type al. with et model for used values Parameter 1 Table day day day / / · · consumer consumer consumer consumer atra est hudb iheog o h Gpeao oehbtnon-significantly exhibit to predator IG the for enough high be should density bacterial · · · consumer consumer consumer P arameter · · / / resource resource resource resource ) ) ) ) ) ) ) m s ) ) 0.3 N.A. 0.7 0.3 N.A. 1 50 2.5 model I Type Chang 1 0.75 1 N.A. 1 × 10 tal.’s et 5 24 1.25 8.14 1.72 values) literature (with model I Type al.’s et Chang 0.1 0.96 0.4 N.A. 0.39 0.11 N.A. 1.25 0.11 N.A. Value × 10 5† .5Lyor n twr (1975) Stewart and Laybourn 1.25 8.14 1.72 model II Type 0.1 0.96 0.4 0.36 0.39 0.11 0.8 (1975) Stewart and Laybourn 1.25 0.11 0.08 × × × 10 10 10 5‡ 5† 5 miial measured Empirically (2011) al. et Fedrigo (1982), al. et Ratkowsky miial measured Empirically 3b Fig. 3b Fig. measured Empirically measured Empirically S2 Appendix see but Colpidium as Same (1980) Fenchel (1975), Stewart and Laybourn S2 Appendix see but Colpidium as Same (1975) Stewart and Laybourn yeI model II Type orefor Source 519 518 517 516 eteivreo adn ie fCliimws1 ie ihrta hto Blepharisma of that than higher times 10 was 1980). Colpidium (Fenchel, of time) handing of inverse the be differen This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. ‡ Accepted Article should (which rate uptake food maximum the that fact the from estimated was value This atra osmto aeaogIPtreatments. IGP among rate consumption bacterial t 25 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 522 521 520 ae .sosteTp Ifntoa epneo nr-ul rdto,ie ubro IG of number i.e. predation, intra-guild of response functional II Type the shows a. Panel 3 which Figure of values, sets different (see text). with line parameterized data dotted the but the fit model and poorly al.’s line et dashed Chang long and from the line) both with both are (long-dashed model) that I Note 2019 (Type literature. model al.’s al. from et values et Chang parameter Chang re-parameterizing with itsstandard when of prediction model line model the the data (solid (3) from the predictions to the differentlines fits (2) three best The error), that 0.01). 100% function and quadratic (p< 80% the The (1) represent strength density mean. the bacteria IGP of higher different error significantly in standard have 468) treatments the represent to bars 298 error state hour The steady (roughly treatments. reached density system protozoa experimental to the respective when with bacteria of density population Mean 2 Figure square rightest the in allowed. line is long-dashed IGP The strength. 250 IGP the 80% represents and 60, 40, four 20, central predator, the of 20 lines the dashed of The thatIG location IGP). left indicates (0% the 20 predator the IG represent from the squares square to first the accessible in not (Colpedium) line the is prey solid in IG prey The C and study. (Blepharisma) and six this predator B in the IG The used represent the species strength). squares of IGP initial rounded 100% the the and are right, square 80 the rounded 60, On 40, 20, left. (0, the treatments andtheexperimen- left on experimental study the this on of design webstructure tal food the simplified showing figure Conceptual 1 Figure Figure This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article captions µ µ ehmnpltstepretg fI ryta sacsil oIG to accessible is that prey IG of percentage the manipulates mesh m ehta sprebet ohI ryadpeao ota 100% that so predator and prey IG both to permeable is that mesh m µ ehta sprebet Gpe u o Gpeao.The predator. IG not but prey IG to permeable is that mesh m 26 565 564 563 562 561 560 559 558 557 556 555 554 553 552 551 550 549 548 547 546 74%o h ainears h G teghtetet 0-0% o Gpeao,IG predator, IG respectively. for density (0%-100%) bacteria treatments and and strength 68.08%, prey IGP model 66%, 6 explain the the lines across from solid variance predictions The the of the model). 67.45% represent II (Type lines barsrepresent response solid error functional The II The Type with mean. the treatments. of strength error IGP standard different density the in protozoa 468) (panel to to respective predator 298 with IG state hour and steady (roughly b) reached (panel system prey experimental IG the when a), c), (panel bacteria of density population Mean 4 Figure resource of is40%. degree the efficiency when assimilation square the of and sum 0.96 residual is total partitioning lowest null the (a has treatments model predation total intra-guild The normalized across model). average higher the in just than results squared model of tile. sum the the residual that of for color combination the model the by the represents represented of and space calculated square White is of density agiven combination sum predator At residual and prey total IG normalized bacteria, toIGpredator. the values, prey parameter IG two the from of ofthe efficiency predator assimilation and combinations the prey different as IG with between well partitioning as model resource the Panel of of degree respectively. square the 0.36 parameters, of unknown and sum two (IG 0.39 residual Blepharisma total be the to of shows estimated time are b. handling prey), and (IG rate Colpidium attack on IGP predator) response, functional II Type the prey This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article osmdaantI ryvru Gpeao ai.Tetoprmtr describing parameters two The ratio. predator IG versus prey IG against consumed 27 566 Figures This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article Figure ( B IG predator ( lepharisma Serratia marcescens, Bacteria consortia Bacillus subtilisBacillus Bacillus cereus,Bacillus 1 ) ( C olpidium IG prey ) ) 0% B C B 20% C 28 + IGP strength (availabilityof IGprey) C B 40% C + C B 60% C + C B 80% C + C B B 100% + + C C This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. Accepted Article Bacteria density at the steady state 0.5 1.0 1.5 2.0 3.5 5.0 Figure 0% ● 2 20% ● Availability ofIGprey (IGPstrength) 40% ● 29 60% ● error barsrepresentstandardofthemean 80%* ● 100%* ● This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. a. Accepted Articleb.

Degree of Number of IG prey consumed Figure resource partitioning (ind./ mL ⋅ day ) 0.94 0.96 0.98 1.0 1.5 2.0 3 0.0 ● ● ● 0.1 ● 10 ● ● Assimilation efficiencyofIGprey IG prey densityperIGpredator(ind./ml) 0.2 ● ● ● 30 ● 0.3 20 ● ● 0.4 ● ● 30 0.5 sum ofsquared total residual Normalized 1.00 1.25 1.50 1.75 2.00 ● This article is protected by copyright. All rights reserved. This article isprotected rights by All copyright. reserved. c. a. Accepted Articleb.

IG predator density IG prey density Bacteria density Figure at the steady state at the steady state at the steady state (ratio to 0% accessibility ) 0.4 0.8 1.2 1.6 1.2 1.6 0.4 0.8 1.2 1.6 0.4 0.8 4 0% 0% 0% 20% 20% 20% Availability ofIGprey (IGPstrength) 40% 40% 40% 31 60% 60% 60% error barsrepresentstandardofthemean 80% 80% 80% 100% 100% 100%