Hydrologic pulsing promotes spatial connectivity and food...

MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol 567 17ndash28 2017httpsdoiorg103354meps12060

Published March 13

INTRODUCTION

Food web dynamics are influenced not only by consumerminusresource interactions but also by environ-mental variation that directly affects ecosystem pro-ductivity habitat suitability and other factors affectingorganism fitness and population growth (Wine milleramp Layman 2005) Further complicating our ability to

predict dynamics in response to environmental driversis the fact that perceived patterns of variation are in-fluenced by temporal and spatial scales of analysis(Petraitis amp Latham 1999) Most and probably all eco-systems are subject to abiotic factors that pulse Re-source pulses in particular are defined as discretesignificant increases in resource availability over in-tervals of time andor space (Yang et al 2008) Envi-

copy Inter-Research 2017 middot wwwint-rescomCorresponding author amgarciaictiofurggmailcom

Hydrologic pulsing promotes spatial connectivity andfood web subsidies in a subtropical coastal ecosystem

A M Garcia1 K O Winemiller2 D J Hoeinghaus3 M C Claudino1 R Bastos1 F Correa1 S Huckembeck1 J Vieira1 D Loebmann1 P Abreu1 C Ducatti4

1Oceanography Institute Rio Grande Federal University Rio Grande Rio Grande do Sul 96203-900 Brazil2Department of Wildlife and Fisheries Sciences and Interdisciplinary Program of Ecology and Evolutionary Biology

Texas AampM University College Station Texas 77843-2258 USA3Department of Biological Sciences and the Advanced Environmental Research Institute University of North Texas Denton

Texas 76203-5017 USA4Paulista Juacutelio de Mesquita Filho State University Stable Isotope Center for Environmental and Life Sciences

Biosciences Institute Botucatu Satildeo Paulo 18608-000 Brazil

ABSTRACT Resource pulsing is a widespread phenomenon but its effects on ecosystem dynam-ics are often difficult to predict Hydrological pulsing in particular is known to influence thestructure and dynamics of fluvial and coastal ecosystems but little information is available aboutits effects on trophic connectivity between wetlands and estuaries We investigated the hypothesisthat hydrologic pulsing drives 1-way trophic subsidies (eg suspended organic matter and fresh-water fish) from wetland to estuary Our study system is a coastal lagoon with an ephemeral mouththat when closed stores freshwater as a sustained flood pulse that is subsequently released whena connection with the sea is reestablished We monitored isotopic composition of consumers andfood sources over the course of an entire flood pulse to infer trophic linkages and spatial subsidiesBefore the flood peak (April and May) freshwater and estuarine zones were largely dependent onlocal primary production sources (seston and C3 plants vs C4 plants and microphytobenthosrespectively) essentially functioning as disconnected compartments A sustained pulse of fresh-water inflow (June to August) induced greater habitat connectivity and a net flow of biomass andenergy from the freshwater zone into the estuarine zone The opening of the lagoon outlet channelabruptly terminated the flood pulse and reduced freshwater subsidies to estuarine consumers andboth zones returned to dependence on autochthonous production Our findings contribute to cur-rent concerns that artificial opening of sandbars in coastal lagoons alters natural ecologicaldynamics with significant effects on biodiversity and ecosystem processes

KEY WORDS Basal resource middot Bayesian mixing model middot Biomass assimilation middot Estuary middot Hydrologic connectivity middot Production source middot Salinity middot Trophic ecology

Resale or republication not permitted without written consent of the publisher

This authors personal copy may not be publicly or systematically copied or distributed or posted on the Open Web except with written permission of the copyright holder(s) It may be distributed to interested individuals on request

Mar Ecol Prog Ser 567 17ndash28 201718

ronmental seasonality affects local population dyna -mics and species interactions but also can have re-gional effects when pulsing factors induce mass mi-grations or synchrony among regio nal populations(Liebhold et al 2004) For example Krenz et al (2011)showed how coastal upwelling influenced by theEl Nintildeo Southern Oscillation affects food web subsi-dies within intertidal ecosystems over a large region Resource subsidies associated with pulsing environ-mental factors strongly influence bottom-up (eg afertilizer effect) and top-down (consumer control) effects in food webs of arid islands (Polis et al 1997Spiller et al 2010)

Pulsing can induce responses that reverberatethroughout the food web (Yang et al 2008) Massspawning by corals and reef fishes in response tolunar and oceanographic cues produces huge subsi-dies for planktivores (Pratchett et al 2001) whichcan have long-term effects on fitness of the latter(McCormick 2003) Regional insect outbreaks inresponse to favorable climatic conditions can devas-tate plant biomass (Mattson amp Addy 1975) while sub-sidizing insectivores (Yang 2004) Seasonal migra-tions can create food web subsidies at the landscapescale (Polis et al 2004) For example the decompos-ing carcasses of post-spawn anadromous salmon pro-vide a pulse of marine-derived nutrients that stimu-lates primary production not only within oligotrophicstreams but also riparian terrestrial ecosystems (Nai -man et al 2002) Theoretical models indicate diverseresponses to pulsing that range from transient popu-lation dynamics that may or may not influence spe-cies coexistence to tipping the dynamics governingalternative stable states of ecosystems (Ostfeld ampKeesing 2000 Holt 2008 Petersen et al 2008)

Hydrologic changes in particular can trigger eco-system pulsing that affects the structure and dyna -mics of fluvial and coastal ecosystems (Winemiller1990 Poff et al 1997 DeAngelis et al 2005 Childers2006 Warfe et al 2011) Fish dynamics in hydrologi-cally pulsed wetland ecosystems monitored over longtime scales (eg Floridarsquos Everglades) are stronglyassociated with seasonal cycles in precipitation andwater depth that cascade to other components of thefood web (DeAngelis et al 1997 Childers 2006)

In southeastern Brazil inter-annual and seasonalvariation in precipitation drives changes in hydro -logy and salinity gradients that in turn affect thecomposition of local fish assemblages within coastallagoons (Garcia et al 2003 2004) A suitable ecosys-tem model for evaluating hydrologic pulsing effectson spatial food web subsidies between freshwaterand estuarine systems in this region is the Lagoa do

Peixe National Park (LPNP) Ecological conditionswithin LPNP are influenced not only by intra-annualvariations in freshwater inflow but also by ephe -meral connections with the sea Freshwater drainingfrom the coastal plain is stored within the lagoon as asustained flood pulse and then released over a rela-tively short period when a connection with the sea isestablished either (1) naturally by inflows or (2) artifi-cially by mechanical excavation (Laneacutes et al 2015)Thus hydrology induces a freshwater flood pulsethat causes aquatic ecosystem expansion with greaterconnectivity between the freshwater wetland and theestuarine zone and conditions promoting input offreshwater basal production sources and consumertaxa into the estuary

We hypothesized that hydrologic pulsing influencesfood web dynamics along the longitudinal fluvial gra-dient via 1-way trophic subsidies (eg passive trans-port of suspended organic matter and dispersal offreshwater fish) from wetland to estuary We used sta-ble isotope methods to test this hypothesis based onextensive temporal and spatial sampling of productionsources and consumers at locations along the longitu-dinal fluvial gradient and used a Bayesian mixingmodel to estimate probability distributions for assimi-lation of alternative production sources Stable isotopeanalysis (SIA) is commonly used to reconstruct trophiclinks connecting food web components across spatial(Hoeinghaus et al 2011 Claudino et al 2015) andtemporal (Claudino et al 2013 Garcia et al 2017)scales and several authors have used SIA to test ef-fects of hydrologic changes on food web organizationand spatial subsidies among ecosystems (Jardine etal 2012 Abrantes et al 2014 Kaymak et al 2015 Ouamp Winemiller 2016) In contrast with most previousstudies that used seasonal or before-and-after sam-pling designs our study analyzed a time series ofsamples taken over frequent short intervals whichenabled us to view food web structure like a movierather than a snapshot This analysis provided a basisfor estimating the timing and magnitude of a spatialfood web subsidy to estuarine consumers

MATERIALS AND METHODS

Study area

LPNP has an area of 344 km2 and is situated alongthe coastal plain of Rio Grande do Sul the southern-most state in Brazil This coastal plain is character-ized by a flat topography low altitude (lt20 m asl)and low tidal range (~05 m Fig 1) Lagoa do Peixe is

Aut

hor c

opy

Garcia et al Hydrologic-driven subsidies in a coastal ecosystem

a shallow coastal lagoon (lt50 cm except its channel)surrounded by freshwater wetlands (Maltchik et al2010) except for its eastern border where sand dunesare prevalent (Fig 1) The lagoon has a narrow outletto the sea that is blocked from FebruaryMarch toAugustSeptember each year by sand dunes untilfreshwater inflows usually from winter rainfall estab-lish a connection If the seasonal connection with thesea is not established naturally earth-moving machin-ery is used to construct an outlet channel (200 m long40 m wide 15 m deep) during late winter (AugustSeptember) The overall dimensions of both artificialand natural channels are similar This periodic open-ing of the outlet is done to promote the entrance ofmarine shrimp larvae into the lagoon to favor com-mercial fisheries and also to drain water from thefloodplain to increase pasture for livestock ranching(Laneacutes et al 2015)

LPNP has a humid subtropical climate with meantemperatures ranging from 146 to 222degC and amean annual temperature of 175degC precipitation inthe study area ranges from 1150 to 1450 mm yrminus1with a mean of 1250 mm yrminus1 Prior studies on estuar-ine systems approximately 100 km south of LPNPhave shown that rainfall anomalies cause hydrologi-cal changes that affect estuarine and freshwater fishassemblages (Garcia et al 2004) To assess the poten-tial influence of climatic variation on ecological dy -namics temperature and rainfall data were ob tainedfrom 2 meteorological stations (Rio Grande andMostardas) near the study area (Fig 1) and the sta-tus of the lagoon connection with the sea was moni-

tored throughout the study Rainfall varied season-ally with 2 periods of high rainfall May to August2008 and January to March 2009 The former wetperiod coincided with a closed lagoon mouth whichoccurred from April to August 2008 At the estuarinesite in particular water flooded over the marginalareas during wet periods increasing hydrologicalexchanges with freshwater wetlands Average airtemperature varied seasonally with lowest valuesduring the austral winter (141degC in June) and high-est values in summer (237degC in February) Based onthis climatic variation we defined flooded and non-flooded phases of the hydrologic pulse in order toevaluate our initial hypothesis (Fig 2)

Field collections and sample processing

Samples were obtained monthly from April 2008 toMay 2009 (with the exception of September 2008 andMarch 2009) in 3 regions along the main longitudinalaxis of the LPNP (1) a freshwater wetland locatednear the northern limit of the park (2) the upperminusmiddle reach of the lagoon that encompasses an eco-tone between estuarine and freshwater zones and(3) the middle reach of the lagoon near the ephe -meral connection with the sea (Fig 1) These sites aresubsequently referred to as freshwater wetlandestuarine zone and lagoon mouth respectively

Fish were caught using 4 fishing gears (gillnetsbeach seine beam trawl dip net) during each monthlysurvey To catch larger fishes (gt200 mm total length)

19

Fig 1 Lagoa do Peixe National Park (LPNP) in the coastal plain of southern Brazil with the location of the sampled areas(squares) at the freshwater wetland (1) at the estuarine zone (2) and at the lagoon mouth (3) Red dashed line denotes theboundaries of the national park Photos taken at the mouth of the lagoon on 2 April and 11 November 2008 show the status of itsconnection with the sea (CD) and the flooded (A) and non-flooded conditions (B) in the same location of the estuarine zone

Photos by A M Garcia

Aut

hor c

opy


2 gillnets (4 times 2 m comprising panels with differentmesh sizes of 15 20 30 and 35 mm) were deployedin the channel (20minus25 m deep) One to 2 beach seine(9 m long 24 m high mesh size 13 mm wings and5 mm center) hauls were made in unvegetated near-shore areas Densely vegetated marginal habitatswere sampled with 3 hauls of the beam trawl (mouth= 1 times 1 m mesh = 5 mm) and dip netting for approxi-mately 15 min Beach seine and beam trawl haulsand dip netting were conducted in shallower waters(lt15 m) and were effective in capturing smaller spe-cies (lt50 mm) including those that take refuge instructurally complex habitats Sampling with multiplegear types provides a more representative sample ofthe fish assemblage especially when the ecosystemis composed of multiple habitat types with differingcomplexities and species of various sizes and behav-iors (Rozas amp Minello 1997 Chick et al 2004 Garciaet al 2006) Representative macro invertebrates suchas adult and immature aquatic insects gastropodspolychaete worms shrimps and crabs were collectedat each study site

Samples of microphytobenthos seston (phytoplank-ton and suspended fine particulate organic matter)and leaves from floating emergent and submergedmacrophytes were collected at each site during eachsurvey This material was used for determination ofthe isotopic composition of major basal productionsources Macrophytes were collected by hand sam-ples of microphytobenthos were obtained by care-fully removing the thin upper layer of flocculent orconsolidated biofilm from substrates and sestonsamples were obtained by filtering water through apre-combusted (450degC 4 h) Whatman glass fiber fil-ter (GFF) using a manual pump

Immediately after collection allspecimens were placed on ice fortransport to the laboratory where theywere stored frozen After thawing fishwere weighed (g) measured (mmtotal length TL) and dissected to ex-tract approximately 5 g of dorsal mus-cle tissue for isotopic analysis For fishlt50 mm TL a composite sample wasobtained by combining muscle tissuefrom 5 to 15 conspecifics from thesame site Gastropods shrimps andcrabs were dissected and a sample ofmuscle tissue was extracted for SIADue to their small sizes polychaetesand aquatic insects were processed aswhole specimens Tissue samples fromfish and macroinvertebrates were in-

spected and any significant non-muscle material (egbone scales exoskeleton) was removed before thesamples were rinsed with distilled water placed insterile Petri dishes and dried in an oven at 60degC untilattainment of a constant weight (minimum of 48 h)Dried samples were ground to a fine powder with amortar and pestle and stored in clean Eppendorf tubesSub-samples were weighed to 6minus10 mg pressed intoUltra-Pure tin capsules (Costech) and sent to theCentro de Isoacutetopos Estaacuteveis Universidade EstadualPaulista (UNESP) for measurement of stable isotoperatios (13C12C and 15N14N) Results are reported asparts per thousand (permil) differences from a correspon-ding standard δHX = [(Rsample Rstandard) minus 1] times 1000where X is a particular element (C carbon or N ni-trogen) the superscript H denotes the heavy isotopemass of the given element (13C 15N) and R is the ratioof the heavy isotope to the light isotope for theelement (13C12C 15N14N) Standards were carbon inthe PeeDee Belemnite and molecular nitrogen in airStandard deviations of δ13C and δ15N replicate analy-ses were 014permil and 013permil respectively

Data analysis

For some analyses fishes were pooled into 5 func-tional guilds based on prior classification proposed byGarcia et al (2001) and also considering Myersrsquo (1938)model of primary- and secondary-division freshwaterfishes as follows (1) estuarine resident species thattypically occur and breed within the estuary (2) estu-arine dependents that are marine or freshwaterspawning species found in large numbers within theestuary during certain periods of their life cycle (3)

20

Fig 2 Monthly variation in rainfall (mm bars) and water temperature (degC line) during the study period (April 2008 to May 2009)

Aut

hor c

opy

Garcia et al Hydrologic-driven subsidies in a coastal ecosystem 21

marine vagrants typically inhabiting marine habitatsand rarely occurring within the estuary (4) primaryfreshwater fishes with no tolerance for salinity thatare confined to freshwater and (5) secondary fresh-water fishes with some salinity tolerance that may en-ter brackish water occasionally Also due to consistentdifferences in retention of 13C by plants using theCalvin cycle (C3) and Hatch-Slack cycle (C4) photo-synthetic pathways (Marshall et al 2007) sampledmacrophytes were pooled and analyzed as 2 groupsC3 vs C4 according to the literature (Marshall et al2007) the range of δ13C values observed in this studyand previous research in nearby coastal lagoon eco-systems (Garcia et al 2007 Hoeinghaus et al 2011)C3 plants tend to have significantly lower average car-bon stable isotope ratios (~ minus27permil) than C4 plants(~ minus14permil) mainly due to differences in the enzymesinvolved in carbon fixation during photosynthesis(Marshall et al 2007) Abundance patterns of thesegroups shift along the salinity gradient of the studysite with C3 being the dominant aquatic macrophytesin the freshwater wetland and C4 plants dominant inthe estuarinemarine sites (Knak 2004 Rolon et al2011)

Bi-plots of δ15N and δ13C values of primary produc-ers and organic sources invertebrates and fisheswere used to evaluate patterns of isotopic variationacross spatial and temporal scales Sources of organiccarbon assimilated by consumers are indicated byproximity between carbon isotope ratios (δ13C) ofconsumers and their food sources whereas trophicposition of each consumer is indicated by the relativeposition of their nitrogen isotope ratios (δ15N) in rela-tion to an isotopic baseline (Peterson amp Fry 1987)Analysis of variance (ANOVA) was used to compareaverage δ13C and δ15N values of sources and con-sumers among study sites and for fishes only amongmonths When a significant main effect was ob servedpairwise comparisons of site means were performedusing a Newman-Keuls post hoc procedure Normal-ity and homogeneity of variances were evaluated bythe Kolmogorov-Smirnov and Cochran tests respec-tively (Zar 1984)

To estimate relative contributions of basal produc-tion sources to fish assemblages of the freshwaterwetland estuarine zone and lagoon mouth we em -ployed Bayesian mixing model with 4 basal produc-tion sources as end-members (seston micro phyto -benthos C3 and C4 plants) using data specific to eachsite Mixing models were run using the Stable Iso-tope Analysis in R (SIAR) package in R statistical soft-ware (R Core Team 2012) which employs a Bayesianapproach that incorporates uncertainties as soci ated

with sample variability and trophic en rich ment onthe assumption they are normally distributed (Parnellet al 2010) Each mixing model was fit via a Markovchain Monte Carlo method that generated simula-tions of plausible values of dietary proportions con-sistent with the data using a Dirichlet prior distribu-tion (Parnell et al 2010) Each model was run basedon 500 000 iterations discarding the first 50 000 andconsidering a non-informative prior to guide thedietary proportion simulations Upper and lowercredibility intervals (95 75 and 50) were used todescribe the range of feasible contributions for eachfood source to the consumer (Parnell et al 2010)

In order to quantify food web subsidies occurringdue to movement of carbon sources from the fresh-water wetland into the estuarine zone we ran 5- end-member SIAR mixing models to estimate on amonthly basis the relative contribution of fresh-water- and estuarine-derived sources for estuarineresident fishes (silversides Atherinella brasiliensisand Odonthestes argentinensis) and individuals ofprimary freshwater fish species that were collectedwithin the estuarine zone during the hydrologicpulse Importantly silversides are not piscivorousand thus assimilation of freshwater-derived sourcesby these estuarine-resident taxa is not by consump-tion of freshwater fishes that moved into the estuarinezone We used average values of carbon and nitrogenisotope ratios of seston and C3 plants collected in thewetland and microphytobenthos C4 plants and ses-ton collected in the estuary as representative fresh-water- and estuarine-derived basal production sourcesWe chose fresh water seston and C3 plants (mostlyfloating macrophytes) because these basal sourcesare transported from the freshwater wetland into theestuarine zone during floods which create hydro -logic connectivity between regions Estuarine sourcesincluded in this model were based on the 4-end-member mixing model that identified the primarysources assimilated by the estuarine fish assemblageMixing models were computed assuming the silver-sides are 2 trophic levels above the basal productionsources and we used values of 054 plusmn 123permil and 254plusmn 011permil to correct for trophic fractionation of isotopicratios of carbon and nitrogen respectively (Vanderk-lift amp Ponsard 2003)

RESULTS

In total 684 samples representing 52 fish specieswere obtained in the freshwater wetland (n = 352) estuarine zone (n = 187) and lagoon mouth (n = 145)

Aut

hor c

opy


between April 2008 and May 2009 for analysis of carbon and nitrogen stable isotope ratios (Table S1in the Supplement at www int-res com articlessupplm 567 p017_ supp pdf) Local fish assemblagesshowed a gradual zonation in species compositionacross the freshwaterminusestuarine gradient Primaryand secondary freshwater fishes were restricted tofreshwater wetland and estuarine zones with theonly exception being Jenynsia multidentata (sec-ondary freshwater species) which oc cur red at all 3sites In contrast estuarine re sidents estuarine de-pendents and marine vagrants were captured onlywithin the estuarine zone and the lagoon mouth(Table S1)

At the assemblage scale there was a significanttrend in δ13C and δ15N along the freshwater to estuar-ine gradient with lower values infreshwater (mean plusmn SD δ13C minus263plusmn 286permil δ15N 75 plusmn 15permil) andmuch higher values in the estuar-ine (δ13C minus157 plusmn 34permil δ15N 104plusmn 18permil) and lagoon mouth zones(δ13C minus138 plusmn 24permil δ15N 133 plusmn21permil Fig 3 Table S1) Overallinvertebrates (eg crabs shrimpspolychaete worms) macro phytes(C3 and C4) and other potentialfood re sources (seston microphy-tobenthos) showed similar trendsSpecifically both δ13C and δ15Nvalues significantly in creased fromthe freshwater wetland to thelagoon mouth (Fig 3 Table S1)

For fish quantitative estimatesfrom Bayesian isotopic mixingmodels revealed that seston and C3

plants (053 to 069 and 023 to044 respectively) were the basalsources assimilated in the greatestproportions by fish inhabiting thefreshwater wetland Converselywithin the 2 mixohaline sites (estu-arine zone and lagoon mouth)C4 plants and microphytobenthoswere the basal carbon sources as -simi lated in the greatest propor-tions For the estuarine zone inparticular C4 plants had a muchgreater contribution to fish bio -mass (070 to 083) comparedwith C3 plants (008 to 021) andC4 plants and microphytobenthoswere the basal carbon sources with

the highest contributions near the lagoon mouth (053to 071 and 007 to 042 respectively Fig 3)

Monthly analysis of the isotopic composition of fishassemblages revealed important changes in the spa-tial trend in average δ13C across study sites (Fig 4)which coincided with the effect that the closure of thelagoonrsquos connection with the sea has on the flood dy-namics of this system During the flooding episoderecorded between April and October 2008 mean fishδ13C values across all sites were significantly lower(minus235permil) compared with those recorded during thenon-flooded period (minus200permil F1687 = 3228 p lt 000)When functional guilds were taken into account weobserved that this difference was primarily due to theoccurrence of salinity intolerant primary freshwaterfishes especially small characids (Astyanax eigen-

22

Fig 3 Mean (+SD) of δ15N and δ13C values for primary producers and organicsources (open circles numbered 1 to 4) prey and fishes (left column) and relativecontributions of basal food sources (C3 and C4 plants seston [Se] and microphyto-benthos [Mi]) to fishes at the freshwater wetland estuarine zone and lagoonmouth sites (right column) Each lsquoDietary proportionrsquo plot shows 50 (gray) 75 (light

gray) and 95 (white) Bayesian credibility intervals

Aut

hor c

opy


manniorum 597 plusmn 118 mm TL A fasciatus 357 plusmn88 mm TL Cheirodon interruptus 324 plusmn 60 mm TL)and poeciliids (Phalloceros caudimaculatus 286 plusmn 87mm TL) that occurred in the estuarine zone from Julyto November (Fig 4) Fish collected from the estuaryin July had lower δ13C values (minus216permil) but the iso-topic composition of these primarily freshwater spe-cies increased (F438 = 751 p lt 000) nearer to the es-tuarine site until approaching values (minus177permil) similarto those observed for estuarine de pendent fishes Es-tuarine resident fishes also showed significant differ-

ences in their average δ13C values during the floodperiod (F1106 = 10837 p lt 000) and had lower values(minus187permil) when compared with the non-flooded pe-riod (minus135permil Fig 4) Temporal variation in δ13C dur-ing flooded vs non-flooded conditions (Fig S1 in theSupplement) also occurred for some basal sourcessuch as microphytobenthos (F118 = 647 p lt 002) andseston (F117 = 1144 p lt 000) which had lower δ13Cduring flooded (minus218 and minus226permil respectively) vsnon-flooded (minus172 and minus184permil respectively) condi-tions Average δ13C values of other sources (C3 and C4

23

Fig 4 Monthly variation in the δ15N and δ13C values for fish guilds (primary freshwater open circles secondary freshwateropen triangles estuarine dependent grey circles estuarine resident grey triangles marine vagrants dark squares) from thefreshwater wetland estuarine zone and lagoon mouth Vertical bar on the right denotes flood status within the estuarine zone

Aut

hor c

opy


plants) did not change significantly between floodingconditions (F116 = 00656 p lt 080 and F15 = 00057p lt 094 respectively Fig S1)

Mixing models revealed that freshwater-derivedbasal sources (C3 plants and seston) contributed toestuarine residents (silversides Atherinella brasilien-sis and Odonthestes argentinensis) during flood con-ditions (April through July Fig 5) Contributions ofthese sources were nearly 0 in April but graduallyincreased to up to 40 in June and July Over the

same period contributions of estuarine-derived basalsources to estuarine fish biomass declined However2 mo after the opening of the lagoon mouth and ces-sation of flood conditions and hydrologic connectivitywith the freshwater wetland the contribution offreshwater-derived basal sources to estuarine resi-dent fish decreased and C4 plants and microphyto-benthos again were the main sources supportingestuarine resident fishes (Fig 5) Primary freshwaterfishes such as characids and poeciliids assimilated

Fig 5 Monthly relative contributions of basal food sources (C3 and C4 plants seston [Sef and See] and microphytobenthos[Mi]) to biomass of estuarine-resident fish (Atherinella brasiliensis Odonthestes argentinensis) (left column) and primaryfreshwater fish (Astyanax eigenmanniorum A fasciatus Cheirodon interruptus Phalloceros caudimaculatus) (right column)from the estuarine zone during flooding and non-flooding conditions Each plot shows the 50 (black) 75 (gray) and 95

(lighter gray) Bayesian credibility intervals of the feasible contributions

Aut

hor c

opy


carbon derived mostly from seston and C3 plants inthe freshwater wetland Conspecifics caught withinthe estuarine zone during the late flooding period(July to August) assimilated freshwater- and estuar-ine-derived basal sources in similar amounts Similarto estuarine resident fishes the percent contributionof estuarine-derived sources (microphytobenthosand C4 plants) to these freshwater species inhabitingthe estuarine zone increased to up to 69 followingcessation of flood conditions (Fig 5)

DISCUSSION

Pulsing hydrology in Lagoa do Peixe affected bothhabitat productivity and habitat connectivity whichtogether resulted in trophic subsidies from the fresh-water wetland to estuarine consumers This subsidyoccurred via passive transport of basal sources (egterrestrial detritus in the form of fine particulateorganic matter) as well as dispersal by freshwaterfishes and probably other freshwater taxa along thefluvial gradient Seston (phytoplankton plus sus-pended particulate organic matter of undeterminedorigin) from the freshwater wetland was apparentlytransported in significant quantities into the estuaryduring the flood pulse The assimilation of this iso-topically distinct seston (approximately 4 and 6permillower δ13C than seston from the estuarine zone andlagoon mouth respectively) by estuarine primaryconsumers (eg zooplankton) and their subsequentconsumption by fish could explain why we observedan average reduction of approximately 5permil in carbonisotope ratios of estuarine fishes (Odonthestes argen-tinesis and Atherinella brasiliensis) during the floodperiod Carbon isotope ratios of estuarine fish gradu-ally shifted back to typical estuarine isotopic valuesobserved during pre-flooding conditions suggestingtheir prey transitioned back to assimilating locallyproduced estuarine seston

In addition to passive transport of seston duringflooding conditions food web subsidies also occurredvia animal dispersal from the wetland into the estu-ary Animal dispersal in fluvial systems is often direc-tional either by downstream drift or upstream swim-ming against the prevailing current (Flecker et al2010 Oliveira et al 2014) For example anadromousfish that ascend rivers to spawn deliver large amountsof nutrients from excretion and carcass de compositionthat subsidize freshwater and riparian ecosystems byenhancing primary production (Nai man et al 2002Koshino et al 2013) Juvenile marine fish thatmigrate into small coastal streams in Brazil excrete

marine-derived nutrients and are consumed by resi-dent freshwater piscivores (Oliveira et al 2014)thereby subsidizing the base and top of the food websimultaneously During the Lagoa do Peixe floodpulse salinity-intolerant primary freshwater fishes(eg small characins such as Astyanax eigenmannio-rum A fasciatus and Cheirodon interruptus) movedinto the estuarine region and began to assimilatematerial derived from local basal production sources

In this system these species generally consumecladocerans other microcrustaceans algae and vas-cular plant fragments (F Correa unpubl data) Con-sumption of locally produced microcrustaceans andvegetation would explain the spatial and temporaldynamics in isotopic values for these species espe-cially for those specimens sampled after the openingof the lagoon mouth In addition to assimilating sig-nificant amounts of estuarine resources during theirbrief periods of residence in the estuary these fresh-water fishes contribute to nutrient dynamics viaexcretion and also delivered a pulse of biomass andenergy that can be exploited by avian (Bugoni et al2005) and mammalian (Colares amp Waldemarin 2000)predators Studies of other coastal lagoons have de -monstrated that a great diversity of freshwater fishesfrom different trophic guilds (eg detritivores zoo -benthi vores piscivores) can colonize estuarine zonesduring high freshwater discharges (Garcia et al2003) However their potential effects on fitnessfeeding habits and survival of estuarine competitorsand predators as well as overall effects to estuarinefood web organization remain largely unknown

Estimates of spatial food web subsidies generallyhave involved asymmetric productivity across aquaticminusterrestrial or marineminusterrestrial ecotones with the moreproductive habitat being a net donor (eg coastalmarine systems riparian forests) and less productivehabitat being a net recipient (eg arid islands foreststreams) (Polis et al 2004) We are unaware of priorevidence for reciprocal subsidies between aquatichabitats with similar productivity such as freshwaterwetlands and estuaries Whether inputs of wetland-derived allochthonous organic matter and freshwaterfish into estuaries induce trophic cascades probablydepends on the relative productivity of the 2 systemsand the degree to which there are reciprocal subsi-dies Potential bottom-up and top-down effects couldbe further elucidated by field experiments For exam-ple Nakano et al (1999) and Nakano amp Mura kami(2001) conducted field ex peri ments and dietary ana -lyses to estimate reciprocal subsidies across aquaticminusterrestrial ecotones and their influence on trophiccascades Baxter et al (2004) demonstrated how habi-

25A

utho

r cop

y


tat alteration and introduction of alien predators sup-pressed flows of resources across ecotones and asso-ciated reciprocal subsidies

Fish guilds responded differently to the pulse offreshwater inflow and degree of connectivity withthe sea Estuarine dependent (eg Brevoortia pecti-nata Micropogonias furnieri Paralichthys orbi gny -anus) and marine vagrant (eg Diapterus rhombeusEucinostomus melanopterus Menticirrhus littoralisTrachinotus marginatus Ulaema lefroyi) fishes wereabsent in mixohaline zones (estuary lagoon mouth)during the latter stages of flooding when characidspoeciliids and other freshwater fishes were preva-lent Marine fishes responded to the freshwater pulseby moving southward toward the sea but they couldnot exit the lagoon because sand dunes blocked theoutlet Several species from the marine guild (egcroakers mullets silversides) were observed in highdensities within shallow pools along the margins ofthe outlet zone Mass mortality of marine speciessometimes occurs during periods of drought as wellas periods of freshwater expansion into mixohalinezones (A M Garcia pers obs) and would explain theabsence of estuarine dependent and marine vagrantfish when the lagoon mouth was closed Similar toother coastal lagoons in the region (Garcia et al 2012)marine fishes returned to the mixohaline zones soonafter the connection to the sea was opened withmachinery Because we did not sample marine habi-tat outside the lagoon outlet we can only speculateabout effects of estuarine discharge of resources onthe nearshore marine food web Using SIA Savage etal (2012) showed that freshwater primary productionsubsidizes suspension-feeding bi valves in coastalwaters of New Zealand when high flow pulses trans-port material from the estuary to the adjacent coast

High-resolution temporal sampling of food webcomponents was crucial to reveal how pulsed fresh-water inflows affect trophic subsidies between zonesof a coastal lagoon ecosystem All estuarine systemsexperience hydrologic pulsing with patterns varyingfrom seasonal (eg wetdry tropics temperate regi onswith spring snowmelt) to highly variable and re la tivelyunpredictable regimes (dryland rivers) Ap pro priatesampling designs are required to reveal spatio temporalpatterns of food web dynamics caused by hydrologicpulsing (Yang et al 2008) Re source pulses occur atmultiple scales and a local subsidy for one species(eg short-lived sedentary organism) may be insignifi-cant for another (eg long-lived migratory organism)(Zackrisson et al 1999 Yang et al 2008) Therefore inorder to quantify resource pulses it is crucial to definerelevant spatial and temporal scales matching re-

sourceminusconsumer inter actions Because we sampledfood web components with greater frequency (monthly)compared with most prior studies (eg pre- vs post-flood or seasonal sampling) our analysis was able toreveal the timing magnitude and duration of trophicsubsidies along the longitudinal fluvial gradient Ana -lysis of fish guilds also allowed us to identify pulses atecologically relevant scales in space and time

Our findings provide evidence that human inter -ference with the connectivity of coastal wetlandslagoons and the sea can greatly affect food web dy -namics including trophic subsidies among their com-partments Mechanical opening of the lagoon outletterminated the flood pulse and reduced delivery offreshwater subsidies to the estuarine zone but it alsoallowed estuarine dependent species to access criti-cal habitat for early life stages Excavation of theLagoa do Peixe outlet channel has been done multi-ple times over the past 150 yr to allow entrance ofshrimp Farfantepenaeus paulensis from the sea intothe lagoon and to drain water from the floodplain toin crease pasture for cattle (Laneacutes et al 2015) Eco -logical responses to artificial connections betweencoastal lagoons and marine systems are difficult topredict and fish species diversity may decline (Saadet al 2002) increase (Griffiths 1999) or change little(Laneacutes et al 2015) There is general concern that arti-ficial opening of sandbars in coastal lagoons altersnatural ecological dynamics with potential negativeeffects on biodiversity and ecosystem services (Grif-fiths 1999 Saad et al 2002) Our study contributes tothis discourse by demonstrating how the timing offreshwater inflows and opening of the outlet affectfood web dynamics in different regions of the lagoonsystem

Future studies of coastal ecotones could examinethe influence of pulsing allochthonous inputs onthe distribution of trophic interactions of varyingstrength within the regional food web Prior theoreti-cal and empirical work suggests that greater propor-tions of weak trophic interactions enhance stabilityin species-rich food webs and strong trophic inter -actions promote destabilizing oscillations (Kokkoriset al 2002 Cross et al 2013) Landscape mosaics in -cluding fluvial ecosystems and other ecotones mighthave greater proportions of weak links within theregional food web when there is periodic ex changeof organisms among habitat compartments (Bellmoreet al 2015) Research that combines comparativeand experimental approaches to estimate interactionstrength could test this hypothesis and further eluci-date ecological responses to pulsing hydrology andfood web subsidies

Aut

hor c

opy


Acknowledgements This study received financial supportfrom Brazilrsquos Conselho Nacional de DesenvolvimentoCien tiacute fico e Tecnoloacutegico CNPq (Grant No 4829202007-6)and International Foundation of Science IFS (Grant NoA4419-1) AMG acknowledges fellowship support fromCNPq (305888 2012-9) and the ICMBIO for providing per-mits (14523-2 and 14523-4) for sample collections KOWacknowledges support from the US National Science Foun-dation (DEB 1257813 and IGERT 0654377) This is a contri-bution of the Grupo de Anaacutelises de Isoacutetopos Estaacuteveis emAmbientes Aquaacuteticos (GAIACNPq)

LITERATURE CITED

Abrantes KG Barnett A Bouillon S (2014) Stable isotope-based community metrics as a tool to identify patterns infood web structure in east African estuaries Funct Ecol28 270minus282

Baxter CV Fausch KD Murakami M Chapman PL (2004)Fish invasion restructures stream and forest food websby interrupting reciprocal prey subsidies Ecology 85 2656minus2663

Bellmore JR Baxter CV Connolly PJ (2015) Spatial com-plexity reduces interaction strengths in the meta-foodweb of a river floodplain mosaic Ecology 96 274minus283

Bugoni L Cormons TD Boyne AW Hays H (2005) Feedinggrounds daily foraging activities and movements ofcommon terns in southern Brazil determined by radio-telemetry Waterbirds 28 468minus477

Chick JH Ruetz CR III Trexler JC (2004) Spatial scale andabundance patterns of large fish communities in fresh-water marshes of the Florida Everglades Wetlands 24 652minus664

Childers DL (2006) A synthesis of long-term research by theFlorida Coastal Everglades LTER Program Hydrobiolo-gia 569 531minus544

Claudino MC Abreu PC Garcia AM (2013) Stable isotopesreveal temporal and between-habitat changes in trophicpathways in a southwestern Atlantic estuary Mar EcolProg Ser 489 29minus42

Claudino MC Pessanha ALM Arauacutejo FG Garcia AM (2015)Trophic connectivity and basal food sources sustainingtropical aquatic consumers along a mangrove to oceangradient Estuar Coast Shelf Sci 167 45minus55

Colares EP Waldemarin HF (2000) Feeding of the neotropi-cal river otter (Lontra longicaudis) in the coastal region ofthe Rio Grande do Sul State southern Brazil IUCN OtterSpec Group Bull 17 6minus13

Cross WF Baxter CV Rosi-Marshall EJ Hall RO Jr and oth-ers (2013) Food-web dynamics in a large river discontin-uum Ecol Monogr 83 311minus337

DeAngelis DL Loftus WF Trexler JC Ulanowicz RE (1997)Modeling fish dynamics and effects of stress in a hydro-logically pulsed ecosystem J Aquat Ecosyst StressRecovery 6 1minus13

DeAngelis DL Trexler JC Loftus WF (2005) Life history trade-offs and community dynamics of small fishes in a season-ally pulsed wetland Can J Fish Aquat Sci 62 781minus790

Flecker AS Mcintyre PB Moore JW Anderson JT TaylorBW Hall RO (2010) Migratory fishes as material and pro-cess subsidies in riverine ecosystems Am Fish Soc Symp73 559minus592

Garcia AM Vieira JP Winemiller KO (2001) Dynamics of theshallow-water fish assemblage of the Patos Lagoon estu-

ary (Brazil) during cold and warm ENSO episodes J FishBiol 59 1218minus1238

Garcia AM Raseira MB Vieira JP Winemiller KO GrimmAM (2003) Spatiotemporal variation in shallow-waterfreshwater fish distribution and abundance in a largesubtropical coastal lagoon Environ Biol Fishes 68 215minus228

Garcia AM Vieira JP Winemiller KO Grimm AM (2004)Comparison of the 1982ndash1983 and 1997ndash1998 El Nintildeoeffects on the shallow-water fish assemblage of the PatosLagoon estuary (Brazil) Estuaries 27 905minus914

Garcia AM Bemvenuti MA Vieira JP Marques DMLMBurns MDM Moresco A Condini MVL (2006) Checklistcomparison and dominance patterns of the fish fauna atTaim Wetland South Brazil Neotrop Ichthyol 4 261minus268

Garcia AM Hoeinghaus DJ Vieira JP Winemiller KO (2007)Isotopic variation of fishes in freshwater and estuarinezones of a large subtropical coastal lagoon Estuar CoastShelf Sci 73 399minus408

Garcia AM Vieira JP Winemiller KO Moraes LE Paes ET(2012) Factoring scales of spatial and temporal variationin fish abundance in a subtropical estuary Mar Ecol ProgSer 461 121minus135

Garcia AM Claudino MC MontrsquoAlverne R Pereyra PERCopertino M Vieira JP (2017) Temporal variability infood assimilation of basal food sources by an omnivorousfish at Patos Lagoon Estuary revealed by stable isotopes(2010ndash2014) Mar Biol Res 13 98ndash107

Griffiths SP (1999) Consequences of artificially openingcoastal lagoons on their fish assemblages Int J Salt LakeRes 8 307minus327

Hoeinghaus DJ Vieira JP Costa CS Bemvenuti CE Wine-miller KO Garcia AM (2011) Estuary hydrogeomorpho -logy affects carbon sources supporting aquatic consumerswithin and among ecological guilds Hydrobiologia 673 79minus92

Holt RD (2008) Theoretical perspectives on resource pulsesEcology 89 671minus681

Jardine TD Pettit NE Warfe DM Pusey BJ and others(2012) Consumerminusresource coupling in wetminusdry tropicalrivers J Anim Ecol 81 310minus322

Kaymak N Winemiller KO Akin S Altuner Z Polat F Dal T(2015) Stable isotope analysis reveals relative influencesof seasonal hydrologic variation and impoundment onassimilation of primary production sources by fish in theUpper Yesilfirmak River Turkey Hydrobiologia 753 131minus147

Knak RB (2004) Management plan for Lagoa do PeixeNational Park IBAMAFNMAFURGNEMAUFPel Bra -siacutelia (in Portuguese)

Kokkoris GD Jansen VAA Loreau M Troumbis AY (2002)Variability in interaction strength and implications forbiodiversity J Anim Ecol 71 362minus371

Koshino Y Kudo H Kaeriyama M (2013) Stable isotope evi-dence indicates the incorporation into Japanese catch-ments of marine-derived nutrients transported byspawning Pacific salmon Freshw Biol 58 1864minus1877

Krenz C Menge BA Freidenburg TL Lubchenco J Chan FFoley MM Nielsen KJ (2011) Ecological subsidies torocky intertidal communities linear or non-linear changesalong a consistent geographic upwelling transitionJ Exp Mar Biol Ecol 409 361minus370

Laneacutes LEK Rolon AS Stenert C Maltchik L (2015) Theeffects of annual artificial opening of a natural sandbar ofa lagoon system on the fish community J Appl Ichthyol31 321minus327

27A

utho

r cop

y


Liebhold A Koenig WD Bjoslashrnstad ON (2004) Spatial syn-chrony in population dynamics Annu Rev Ecol Syst 35 467minus490

Maltchik L Laneacutes LEK Stenert C Medeiros ESF (2010) Spe-cies-area relationship and environmental predictors offish communities in coastal freshwater wetlands of south-ern Brazil Environ Biol Fishes 88 25minus35

Marshall JD Brooks JR Lajtha K (2007) Sources of varia-tion in the stable isotopic composition of plants In Mitchener R Lajtha K (eds) Stable isotopes in ecologyand environmental science Blackwell Scientific Oxfordp 22minus60

Mattson WJ Addy ND (1975) Phytophagous insects as regu-lators of forest primary production Science 190 515minus522

McCormick MI (2003) Consumption of coral propagules aftermass spawning enhances larval quality of damselfishthrough maternal effects Oecologia 136 37minus45

Myers GS (1938) Freshwater fishes and West Indian zoo-geography Annu Rep Board Regents Smithson Inst 92 339minus364

Naiman RJ Bilby RE Schindler DE Helfied JM (2002) Pacificsalmon nutrients and the dynamics of freshwater andriparian ecosystems Ecosystems 5 399minus417

Nakano S Murakami M (2001) Reciprocal subsidies dyna -mic interdependence between terrestrial and aqua ticfood webs Proc Natl Acad Sci USA 98 166minus170

Nakano S Miyasaka H Kuhara N (1999) Terrestrial-aquaticlinkages riparian arthropod inputs alter trophic cascadesin a stream food web Ecology 80 2435minus2441

Oliveira MCLM Bastos RF Claudino MC Assumpccedilatildeo CMGarcia AM (2014) Transport of marine-derived nutrientsto subtropical freshwater food webs by juvenile mullets a case study in southern Brazil Aquat Biol 20 91minus100

Ostfeld RS Keesing F (2000) Pulsed resources and commu-nity dynamics of consumers in terrestrial ecosystemsTrends Ecol Evol 15 232minus237

Ou C Winemiller KO (2016) Seasonal hydrology shifts pro-duction sources supporting fishes in rivers of the LowerMekong Basin Can J Fish Aquat Sci 73 1342minus1362

Parnell AC Inger R Bearhop S Jackson AL (2010) Sourcepartitioning using stable isotopes coping with too muchvariation PLOS ONE 5 e9672

Petersen JK Hansen JW Laursen MB Clausen P Carsten -sen J Conley DJ (2008) Regime shift in a coastal marineecosystem Ecol Appl 18 497minus510

Peterson BJ Fry B (1987) Stable isotopes in ecosystem stud-ies Annu Rev Ecol Evol Syst 18 293minus320

Petraitis PS Latham RE (1999) The importance of scale intesting the origins of alternative community states Eco -logy 80 429minus442

Poff NL Allan JD Bain MB Karr JR and others (1997) Thenatural flow regime a paradigm for river conservationand restoration Bioscience 47 769minus784

Polis GA Anderson WB Holt RD (1997) Toward an integra-tion of landscape and food web ecology the dynamics ofspatially subsidized food webs Annu Rev Ecol Syst 28 289minus316

Polis GA Power ME Huxel GR (eds) (2004) Food webs at thelandscape level University of Chicago Press Chicago IL

Pratchett MS Gust N Goby G Klanten SO (2001) Consump-tion of coral propagules represents a significant trophiclink between corals and reef fish Coral Reefs 20 13minus17

R Core Team (2012) R a language and environment for sta-tistical computing R Founadation for Statistical Comput-ing Vienna wwwr-projectorg

Rolon AS Rocha O Maltchick L (2011) Diversity of aquaticmacrophytes in the Lagoa do Peixe National ParkNeotrop Biol Conserv 6 5minus12

Rozas LP Minello TJ (1997) Estimating densities of smallfishes and decapod crustaceans in shallow estuarinehabitats a review of sampling design with focus on gearselection Estuaries 20 199minus213

Saad AM Beaumond AC Caramaschi EP (2002) Effects ofartificial canal openings on fish community structure ofImboassica coastal lagoon Rio de Janeiro Brazil J CoastRes 36 634minus639

Savage C Thrush SF Lohrer AM Hewitt JE (2012) Ecosys-tem services transcend boundaries estuaries provideresource subsidies and influence functional diversity incoastal benthic communities PLOS ONE 7 e42708

Spiller DA Piovia-Scott J Wright AN Yang LH Takimoto GSchoener TW Iwata T (2010) Marine subsidies have mul-tiple effects on coastal food webs Ecology 91 1424minus1434

Vanderklift MA Ponsard S (2003) Sources of variation inconsumer-diet δ15N enrichment a meta-analysis Oeco -logia 136 169minus182

Warfe DM Pettit NE Davies PM Pusey BJ and others (2011)The lsquowet-dryrsquo in the wet-dry tropics drives ecosystemstructure and function of northern Australian riversFreshw Biol 56 2169minus2195

Winemiller KO (1990) Spatial and temporal variation in trop-ical fish trophic networks Ecol Monogr 60 331minus367

Winemiller KO Layman CA (2005) Food web science movingon the path from abstraction to prediction In Ruiter PCWolters V Moore JC (eds) Dynamic food webs multi-species assemblages ecosystem development and envi-ronmental change Elsevier Amsterdam p 10minus23

Yang LH (2004) Periodical cicadas as resource pulses inNorth American forests Science 306 1565minus1567

Yang LH Bastow JL Spence KO Wright AN (2008) Whatcan we learn from resource pulses Ecology 89 621minus634

Zackrisson O Nilsson MC Jaderlund A Wardle DA (1999)Nutritional effects of seed fall during mast years in borealforest Oikos 84 17minus26

Zar JH (1984) Biostatistical analysis 2nd edn Prentice-HallEnglewood Cliffs NJ

Editorial responsibility Jana Davis Annapolis Maryland USA

Submitted March 7 2016 Accepted January 17 2017Proofs received from author(s) February 21 2017

Aut

hor c

opy


ronmental seasonality affects local population dyna -mics and species interactions but also can have re-gional effects when pulsing factors induce mass mi-grations or synchrony among regio nal populations(Liebhold et al 2004) For example Krenz et al (2011)showed how coastal upwelling influenced by theEl Nintildeo Southern Oscillation affects food web subsi-dies within intertidal ecosystems over a large region Resource subsidies associated with pulsing environ-mental factors strongly influence bottom-up (eg afertilizer effect) and top-down (consumer control) effects in food webs of arid islands (Polis et al 1997Spiller et al 2010)

Pulsing can induce responses that reverberatethroughout the food web (Yang et al 2008) Massspawning by corals and reef fishes in response tolunar and oceanographic cues produces huge subsi-dies for planktivores (Pratchett et al 2001) whichcan have long-term effects on fitness of the latter(McCormick 2003) Regional insect outbreaks inresponse to favorable climatic conditions can devas-tate plant biomass (Mattson amp Addy 1975) while sub-sidizing insectivores (Yang 2004) Seasonal migra-tions can create food web subsidies at the landscapescale (Polis et al 2004) For example the decompos-ing carcasses of post-spawn anadromous salmon pro-vide a pulse of marine-derived nutrients that stimu-lates primary production not only within oligotrophicstreams but also riparian terrestrial ecosystems (Nai -man et al 2002) Theoretical models indicate diverseresponses to pulsing that range from transient popu-lation dynamics that may or may not influence spe-cies coexistence to tipping the dynamics governingalternative stable states of ecosystems (Ostfeld ampKeesing 2000 Holt 2008 Petersen et al 2008)

Hydrologic changes in particular can trigger eco-system pulsing that affects the structure and dyna -mics of fluvial and coastal ecosystems (Winemiller1990 Poff et al 1997 DeAngelis et al 2005 Childers2006 Warfe et al 2011) Fish dynamics in hydrologi-cally pulsed wetland ecosystems monitored over longtime scales (eg Floridarsquos Everglades) are stronglyassociated with seasonal cycles in precipitation andwater depth that cascade to other components of thefood web (DeAngelis et al 1997 Childers 2006)

In southeastern Brazil inter-annual and seasonalvariation in precipitation drives changes in hydro -logy and salinity gradients that in turn affect thecomposition of local fish assemblages within coastallagoons (Garcia et al 2003 2004) A suitable ecosys-tem model for evaluating hydrologic pulsing effectson spatial food web subsidies between freshwaterand estuarine systems in this region is the Lagoa do

Peixe National Park (LPNP) Ecological conditionswithin LPNP are influenced not only by intra-annualvariations in freshwater inflow but also by ephe -meral connections with the sea Freshwater drainingfrom the coastal plain is stored within the lagoon as asustained flood pulse and then released over a rela-tively short period when a connection with the sea isestablished either (1) naturally by inflows or (2) artifi-cially by mechanical excavation (Laneacutes et al 2015)Thus hydrology induces a freshwater flood pulsethat causes aquatic ecosystem expansion with greaterconnectivity between the freshwater wetland and theestuarine zone and conditions promoting input offreshwater basal production sources and consumertaxa into the estuary

We hypothesized that hydrologic pulsing influencesfood web dynamics along the longitudinal fluvial gra-dient via 1-way trophic subsidies (eg passive trans-port of suspended organic matter and dispersal offreshwater fish) from wetland to estuary We used sta-ble isotope methods to test this hypothesis based onextensive temporal and spatial sampling of productionsources and consumers at locations along the longitu-dinal fluvial gradient and used a Bayesian mixingmodel to estimate probability distributions for assimi-lation of alternative production sources Stable isotopeanalysis (SIA) is commonly used to reconstruct trophiclinks connecting food web components across spatial(Hoeinghaus et al 2011 Claudino et al 2015) andtemporal (Claudino et al 2013 Garcia et al 2017)scales and several authors have used SIA to test ef-fects of hydrologic changes on food web organizationand spatial subsidies among ecosystems (Jardine etal 2012 Abrantes et al 2014 Kaymak et al 2015 Ouamp Winemiller 2016) In contrast with most previousstudies that used seasonal or before-and-after sam-pling designs our study analyzed a time series ofsamples taken over frequent short intervals whichenabled us to view food web structure like a movierather than a snapshot This analysis provided a basisfor estimating the timing and magnitude of a spatialfood web subsidy to estuarine consumers

MATERIALS AND METHODS

Study area

LPNP has an area of 344 km2 and is situated alongthe coastal plain of Rio Grande do Sul the southern-most state in Brazil This coastal plain is character-ized by a flat topography low altitude (lt20 m asl)and low tidal range (~05 m Fig 1) Lagoa do Peixe is

Aut

hor c

opy








19



Aut

hor c

opy






Data analysis


20


Aut

hor c

opy







RESULTS


Aut

hor c

opy








22



Aut

hor c

opy




23


Aut

hor c

opy







Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy








19



Aut

hor c

opy






Data analysis


20


Aut

hor c

opy







RESULTS


Aut

hor c

opy








22



Aut

hor c

opy




23


Aut

hor c

opy







Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy






Data analysis


20


Aut

hor c

opy







RESULTS


Aut

hor c

opy








22



Aut

hor c

opy




23


Aut

hor c

opy







Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy







RESULTS


Aut

hor c

opy








22



Aut

hor c

opy




23


Aut

hor c

opy







Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy








22



Aut

hor c

opy




23


Aut

hor c

opy







Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy




23


Aut

hor c

opy







Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy







Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy



DISCUSSION






25A

utho

r cop

y








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy








Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy



LITERATURE CITED
































27A

utho

r cop

y






































Aut

hor c

opy






































Aut

hor c

opy

Hydrologic pulsing promotes spatial connectivity and food...

Documents

Transcript of Hydrologic pulsing promotes spatial connectivity and food...