Estuarine, Coastal and Shelf Science 69 (2006) 491e504www.elsevier.com/locate/ecss

Effect of rainfall as a component of climate change on estuarinefish production in Queensland, Australia

Jan-Olaf Meynecke a,*, Shing Yip Lee a, Norman C. Duke b, Jan Warnken a

a Centre for Aquatic Processes and Pollution, and School of Environmental and Applied Science, Griffith University - Gold Coast Campus,PMB 50 GCMC, Queensland, 9726, Australia

b Centre for Marine Studies, University of Queensland, St Lucia, Queensland, 4072, Australia

Received 17 February 2006; accepted 8 May 2006

Available online 19 June 2006

Abstract

The speculation that climate change may impact on sustainable fish production suggests a need to understand how these effects influence fishcatch on a broad scale. With a gross annual value of A$ 2.2 billion, the fishing industry is a significant primary industry in Australia. Manycommercially important fish species use estuarine habitats such as mangroves, tidal flats and seagrass beds as nurseries or breeding groundsand have lifecycles correlated to rainfall and temperature patterns. Correlation of catches of mullet (e.g. Mugil cephalus) and barramundi (Latescalcarifer) with rainfall suggests that fisheries may be sensitive to effects of climate change. This work reviews key commercial fish and crus-tacean species and their link to estuaries and climate parameters. A conceptual model demonstrates ecological and biophysical links of estuarinehabitats that influences capture fisheries production. The difficulty involved in explaining the effect of climate change on fisheries arising fromthe lack of ecological knowledge may be overcome by relating climate parameters with long-term fish catch data. Catch per unit effort (CPUE),rainfall, the Southern Oscillation Index (SOI) and catch time series for specific combinations of climate seasons and regions have been exploredand surplus production models applied to Queensland’s commercial fish catch data with the program CLIMPROD. Results indicate that up to30% of Queensland’s total fish catch and up to 80% of the barramundi catch variation for specific regions can be explained by rainfall often witha lagged response to rainfall events. Our approach allows an evaluation of the economic consequences of climate parameters on estuarine fish-eries, thus highlighting the need to develop forecast models and manage estuaries for future climate change impact by adjusting the quota forclimate change sensitive species. Different modelling approaches are discussed with respect to their forecast ability.� 2006 Elsevier Ltd. All rights reserved.

Keywords: estuarine fish catch; climate change; conceptual model; modelling; Australia

1. Introduction

When investigating climate effects on fish stocks, the focushas been on commercially important pelagic fish species suchas tuna (Thunnus albacares, Thunnus thynnus), mackerel(Trachurus declivis, Trachurus novaezelandiae) and sardines(Sardinops sagax, Sardinops pilchardus) (Klyashtorin, 1998;Yanez et al., 2002; Lloret et al., 2004). In particular, therehas been a number of studies demonstrating sensitivity of

* Corresponding author.

E-mail address: j.meynecke@griffith.edu.au (J.-O. Meynecke).

0272-7714/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.ecss.2006.05.011

small pelagic species to El Nino Southern Oscillation(ENSO) fluctuations (Schwartzlose et al., 1999; FAO, 2000;Lea, 2000; Lehodey et al., 2003). Potential climate effectson estuary-dependent fisheries species have been neglectedso far. It is reasonable to expect that on-going climaticchanges may also affect a number of estuary-dependent fishspecies such as mullet (e.g. Mugil cephalus), which is oneof the most important animal protein sources in many regionsof the world (Hill, 2004), and some prawn species (e.g.,Penaeus esculentus and Penaeus plejebus), with significantimplications for sustainable fish production. There are com-mercial fisheries in estuaries and near-shore waters of tropicalAustralia for penaeid prawns (e.g., Penaeus merguiensis,

492 J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

Penaeus indicus, Penaeus esculentus, Penaeus semisulcatus,Metapenaeus ensis), finfish (e.g., Lates calcarifer, Polydacty-lus macrochir, Eleutheronema tetradactylum, Mugil spp.,Liza vaigiensis), sharks (e.g., Carcharhinus tilstoni, Carchar-hinus sorrah) and mud crabs (Scylla serrata), having a com-bined annual value of about A$ 2.2 billion (Dent, 2002;Robins et al., 2005). However, potential economic impactson fish catch caused by climate change are unclear so far(Abbs, 2002). This gap is mainly caused by the difficulty inestablishing a clear link between commercial fish speciesand climate variables on a broad scale. Large variability in cli-mate, the catch and biology of these species make generalisa-tions difficult and the local nature of most studies restricts theusefulness of modelling approaches. It remains an importanttask to develop climatic indices capable of improving fisheriesmanagement (Roessig et al., 2004). Studies in Queensland,Australia (Glaister, 1978; Vance et al., 1998; Loneragan andBunn, 1999; Staunton-Smith et al., 2004; Growns and James,2005) indicated that there is a strong relation between fresh-water runoff and some important commercial fisheries species(e.g., mullet (Mugil spp.), flathead (Platycephalus spp.), whit-ing (Sillago spp.), prawns (Family Penaeidae), mud crabs (S.serrata)) but it remains unclear whether these links also applyto the whole coast of Queensland and if a global climate indi-cator such as the Southern Oscillation Index (SOI) has mea-surable influence on fish catch. The 3000 km of Queenslandcoastline offer a wide range of rainfall, temperature andhigh records of fish catch for a broad scale modelling ap-proach (Fig. 1).

A number of different models have been developed world-wide to study links of fisheries to climate and their economicconsequences. Modified simple surplus production modelsincluding both environmental and bio-economic diversity vari-ables such as the GordoneSchaefer model or the GullandeFoxmodel to relate banana prawn (Penaeus merguiensis) commer-cial catch with rainfall have been used (Kasulo and Perrings,2004). Furthermore, bio-economic models were established

for more detailed analysis where information on stock dynamics(e.g., mortality) was available (Hannesson, 1993; FAO, 1994,1997; King, 1995; Sparre and Venema, 1997; Ulrich et al.,2001). However, if the population dynamics are largely un-known, simple models (e.g., biomass dynamic population modelwith environmental parameters) as available in CLIMPROD(FAO, 1998) can provide useful outcomes for preliminary inves-tigations (Chen et al., 1997; Evans et al., 1997; FAO, 2000).

The main objectives of this paper are to establish whetherthere is any historical association between rainfall, SOI andthe catch of selected Queensland estuary-dependent fisheriesspecies and provide a broad-scale case to model climatic ef-fects on fish landings and potential economic losses

2. Climate change in Australia

The climatic analysis of the Intergovernmental Panel onClimate Change (IPCC) Report for Australia resulted in a trendto greater dryness in the next 50e100 years and a clear in-crease in temperature (IPCC, 2001). In accordance, CSIRO re-gional projections predicted an increase in annual averagetemperatures of 0.4e2.0 �C by 2030 (relative to 1990) and1.0e6.0 �C by 2070. Considerable uncertainty remains withfuture changes in rainfall for 2070, which vary betweenþ10% and �35% for the east coast of Australia (Walshet al., 2000; Hughes, 2003). Such variations are closely linkedto Southern Oscillation fluctuations similar to Northern Pacificatmospheric forcing (Norton and McLain, 1994). Over the lastdecades, El Nino events increased in frequency and intensitywith the SOI corresponding to a rise in global temperatures.The Southern Oscillation Index (SOI) is in general in phasewith Australian rainfall on interannual time scales with thestrongest correlations in southeast Australia (Power et al.,1999). The streamfloweENSO connection is strongest inlate spring and summer in most parts of Australia, while thestreamflow serial correlation is significant for most parts ofthe year, suggesting a clear dependence between ENSO,

Fig. 1. Map showing major river systems, cities and rainfall distribution (source: Geoscience_Australia, 2005).

493J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

rainfall and freshwater runoff (Power et al., 1999). The abilityto forecast El Nino can provide a longer lead-time for devel-oping strategies to deal with expected impactsdincluding eco-nomic impacts on fisheries.

3. Links between climate variables and estuary-dependentfish species

Efforts have been made to understand linkages betweenphysical changes such as El Nino events in the ocean environ-ment and biological processes that influence commercial im-portant fish stocks (Evans et al., 1995; Schwartzlose et al.,1999; Lea, 2000; Byrne et al., 2002; Eide and Heen, 2002;Currie and Small, 2005). In the last 30 years, studies in Aus-tralia and elsewhere, predominantly on prawns and estuary-dependent species, are suggesting a sensitivity of these speciesto freshwater runoff leading to fluctuation in catch (Table 1).Available evidence demonstrated that river flow is a criticalfactor in maintaining nutrient and detrital input to estuaries,as well as preventing the development of hypersaline condi-tions within these systems (Blaber and Blaber, 1980; Robert-son and Duke, 1990; Forbes and Cyrus, 1993; Whitfield,1994). Knowledge of the freshwater flow requirements of fish-eries is based on the analysis of catch (¼ landings) and fresh-water flow data. For example, Loneragan and Bunn (1999)showed that total fish catch corresponds with freshwater runoff(r ¼ 0.72; P < 0.05; n ¼ 8) for the Logan River, Queensland,Australia, particularly for mullet (Mugil spp.) and flathead(Platycephalus spp.). Furthermore, it is well known that higherrainfall in the wet season translates to good prawn seasons.Simple models based on rainfall, monthly run off data andmonthly sea surface temperature (SST) are used for predic-tions in the Gulf of Carpentaria prawn fishery, as emigrationrates of juvenile banana prawns from estuaries to nearshoreareas are strongly linked to rainfall events (Staples and Vance,1986; Vance et al., 1998) and the emigration rates are signif-icantly correlated with commercial catches (Staples andVance, 1986; Vance et al., 1998). Most correlations betweenfreshwater flow, rainfall and prawn catch have been reported

for species with the greatest tolerance or exploitation of brack-ish-water habitats. In general, significant positive relationshipsoccur between annual catch and total rainfall (or freshwaterflow) in the same or previous year (Gunter and Hildebrand,1954; Ruello, 1973; Glaister, 1978; Vance et al., 1985; Gam-melsrod, 1992; Chen et al., 1994; Galindo-Bect et al., 2000).Significant within-year correlations between catch andmonthly or seasonal freshwater flow (or rainfall) have alsobeen reported (Glaister, 1978; Browder, 1985; Vance et al.,1985,1998; Gammelsrod, 1992; Evans et al., 1997). Otherstudies found a significant positive relation between spawningseason, runoff and barramundi fish catch (Staunton-Smithet al., 2004).

Based on the above literature review, we hypothesise thatrainfall modulates catchability and/or vulnerability of estua-rine fish species and affects estuarine fish catch. Therefore,the effect of years and months with extreme rainfall eventsis predicted to be detectable in fish catch data throughoutQueensland. Overall, regional differences in the relationshipare expected as the variability of rainfall throughout Queens-land is high and fish species have probably adapted to thesedifferences, either genetically or in terms of behaviouralchange. Genetically discrete stocks for barramundi (Lates cal-carifer) were found to exist in different groups of river systemsin northern Australian waters (Quinn, 1987; Watts and John-son, 2004) and of mud crabs (Scylla serrata) between theGulf of Carpentaria and the Queensland east coast (Gopurenkoand Hughes, 2002).

Strong links between temperature and fish catch have notbeen reported for tropical Australia so far. Studies in Africanestuaries showed that effects may occur on the level of fish as-semblages (Whitfield and Harrison, 2003; Whitfield, 2005).Underlying mechanisms (e.g., initiation of breeding cycles inmud crabs) may be controlled by trigger values, which are un-likely to be detected in broad spatial and temporal analyses intropical and subtropical marine environments where taxa areless affected by increasing temperature (Smith, 1990). There-fore, we do not expect temperature to have a measurable influ-ence on fish catch data and have focused on rainfall and SOI

Table 1

Studies showing a positive relationship between fish catch species and environmental factors

Author Factor Fisheries species Region

Gunter and Hildebrand (1954) Rainfall White shrimp (Penaeus setiferus) USA, Texas

Ruello (1973), Glaister (1978) Rainfall School prawn (Metapenaeus macleayi) Australia, Hunter River

Vance et al. (1985), Staples and

Vance (1986), Vance et al. (1998)

Rainfall, freshwater

runoff

Banana prawn (Penaeus merguiensis) Australia, Gulf of Carpentaria

Smith (1990) Rainfall, temperature Blue crab (Callinectes sapidus) USA, Louisiana

Gammelsrod (1992) Rainfall Red-legged banana prawn (Penaeus indicus) Mozambique

Loneragan and Bunn (1999) Freshwater runoff Mullet (Mugil spp.); flathead (Platycephalus spp.); tiger

(Penaeus semisulcatus), school (Metapenaeus macleayi),greasy (Metapenaeus bennettae) and king prawns

(Penaeus plejebus); mud crab (Scylla serrata)

Australia, Logan River

Galindo-Bect et al. (2000) Rainfall Blue shrimp (Litopenaeus stylirostris) Mexico, Gulf of Calif.

Browder et al. (2002) Rainfall Pink shrimp (Penaeus duorarum) USA, Gulf of Mexico

Powell et al. (2002) Freshwater runoff Brown shrimp (Penaeus aztecus) USA, Texas

Staunton-Smith et al. (2004) Freshwater runoff Barramundi (Lates calcarifer), School prawns

(Metapenaeus macleayi)

Australia, Fitzroy River

Growns and James (2005) Freshwater runoff Australian bass (Macquaria novemaculeata) Australia, Hawkesbury River

494 J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

for estuarine fish species, particularly, mullet (Mugil spp.), pe-naeid prawns (Family Penaeidae) and barramundi (Lates cal-carifer) in Queensland, Australia. However, monthly averagetemperature as been considered in the analyses for the tworegions.

4. Methods

4.1. Estuary-dependent species and climate: aconceptual model

The relevant processes related to commercial fish catch inestuaries are described in a model to allow better understand-ing of the complex interactions between climate parameter andfor estuary-dependent fish stock. For example, several modelshave been used to conceptualise the role of rainfall in estuarineecosystems. An ‘order of effects’ model, which distinguishesbetween the direct (salinity, erosion) and indirect effects (abi-otic and biotic variables) of flow (rainfall) on ecological pro-cesses is suggested by Hart and Finelli (1999). Kimmerer(2002) proposed a simple conceptual model, where rainfalland with it freshwater inflow to estuaries results in physical,chemical and biological consequences. A similar approachwas given by Alber (2002) who suggested that freshwater in-flow affects estuarine conditions (e.g., physical and chemicalconditions), which in turn affect estuarine resources (e.g.,plants and animals) and ecological processes. Sklar and Brow-der (1998) also included the effects of landscape modifica-tions, tidal actions and solar activity. Following theseapproaches a conceptual model for this study has been devel-oped in which climate-relevant processes and parametersare related to economic consequences in a temporal dimen-sion (Fig. 2). Three categories respectively representing theabiotic factors, the biological response and the economic

consequences are suggested for the model. Starting with an in-crease in atmospheric temperature a number of further abioticfactors such as rainfall experience a significant shift, leading tobiological response of dependent fish species by, for example,actively avoiding the estuary or indirectly by an increasedmortality rate of juveniles causing a recruitment failure (Lone-ragan and Bunn, 1999; Robins et al., 2005) or a change in fishassemblages (Whitfield, 2005). In the long term temperatureincrease may induce a rise in sea level, causing an ecotoneshift of important nursery habitats and resulting in reducedhabitat availability for estuarine fish species. As a consequencefish catch may be reduced and fishing pressure increases (e.g.,increased number of fishing days), risking overexploitationand economic loss. A decline in income for the fishermenmay then lead to conflicts with protection zones and currentmanagement strategies.

4.2. Data

Data on catch, effort (number of days and boats) and grossvalue of production for estuary-dependent species or speciesgroups were provided by the Queensland Department of Pri-mary Industries and Fisheries (QDPI&F) Assessment andMonitoring Unit (Table 2). This dataset is based on daily log-book records reported by commercial fishers, covering theyears 1988e2004. The data has an estimated error of 10%due to the nature of recording, fluctuations by market require-ments, policies and management changes. In addition, avail-able data on mullet catch for the years 1945e1980 wereobtained from QDPI&F reports but have not been pooledwith recent data sets due to major differences in the census.The Queensland fisheries 1988e2000 report (Williams,2002) provided additional data for some species mainly forthe east coast of Queensland. Effort has been calculated for

Atmospheric temperature

Salinity, oxygen, DOC

Precipitation & runoff

Wind speed & direction

Ocean temperature

Sea level

Currents Light

COCO22

Directio

n o

f ch

an

ge

Fish exploitation policies

Industrial strategies

Food security, economic loss Habitat change

Stress, spawning, movements

Change in food supply

Fish assemblages & abundance

Fishing pressure

Directio

n o

f ch

an

ge

Fig. 2. A conceptual model of the complex interaction between estuary-dependent species and climate influenced abiotic factors. The effects are increasing from

bottom to top.

495J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

the individual fisheries (line, trawl, net and pot) and pooledfor each species. Data on annual and monthly rainfall forthe time period 1988e2004 and values for the SOI (1945e2004) were obtained from the Commonwealth Bureau of Me-teorology (BOM) as well as rainfall and temperature datafrom weather stations in Cairns and Moreton Bay. These re-gions have been selected since the weather data are consistentover a long time and an intensive fishing activity within Trin-ity Inlet and Moreton Bay provides high number of catch re-cords. Data from the Barron River gauging station (Cairns,Queensland) on freshwater runoff has been provided by theDepartment of Natural Resources and Mines. This gaugingstation represents a major part of the freshwater runoff intoTrinity Inlet. Other gauging stations were found to have sig-nificant data gaps. For the rainfall, regional station data aswell as monthly average coastal rainfall data was used. Be-cause of the known seasonality, with the maximum rainfallin Queensland occurring in December to February, and theminimum during May and October, the ‘annual’ rainfall,SOI and catch data were grouped into wet (November toApril) and dry seasons (May to October). Furthermore, we di-vided the coast of Queensland into 8 different regions accord-ing to the Bureau of Meteorology rainfall districts (BOM,2004) and used coastal catch data from 13 different speciesgroups (Table 2) to observe for regional differences in the cor-relation between rainfall and fish catch (Fig. 3). These 8 re-gions also happen to reflect some of the differences inspecies composition between coastal regions of Queensland

Table 2

Major species or species group selected for the estuary-dependent fish catch

data. Criteria for species selection: The species should be/have (1) relative

constant and high market value, (2) estuary-dependent, (3) widespread

throughout Queensland. Source: Yerasley et al. (1999)

Common name

and fish catch class

Taxa

Barramundi Lates calcarifer

Bream Nematalosa spp., Monodactylus argenteus,

Pomadasys maculatum, Acanthopagrus australis,

A. berda

Bugs Thenus indicus, T. orientalis

Mud Crab Scylla serrataFlathead e Dusky Platycephalus fuscus

Grunter Pomadasys spp., P. kaakan

Mullet Liza vaigiensis, Valamugil georgii, Mugil cephalus,

L. argenteaPrawns e Bait Family Penaeidae

Prawns e Banana Penaeus merguiensis

Prawns e Bay Metapenaeus bennettae, M. insolitus

Prawns e Endeavour Metapenaeus endeavouri, M. ensisPrawns e Greasy Metapenaeus macleayi

Prawns - King Penaeus plejebus

Prawns e School Metapenaeus macleayi

Prawns e Tiger Penaeus esculentus, P. semisulcatusSea Perch e Mixed Family Lutjanidae

Mangrove Jack Lutjanus argentimaculatus

Threadfin King/Blue Eleutheronema tetradactylum, Polydactylusmacrochir sheridani

Whiting Sillago ciliata, S. analis, S. maculata, S. burrus,

S. ingenuua, S. sibama, S. robusta

(e.g., Gulf of Carpentaria, North Queensland, Central Queens-land and Southern Queensland).

4.3. Data analyses

As a first attempt, El Nino events (1991e1992, 1993e1995, 1997e1998) and related rainfall events have been com-pared with fish catch data from Queensland, Moreton Bay andCairns. Relationships between catch, temperature, rainfall,and SOI were explored using single linear regression modelsand correlation analysis. We used this approach instead ofthe time series models (e.g., transfer function) because the de-pendent variable (catch) in the time series has a maximumlength of 17 years, which is less than the minimum necessary(i.e. approximately 50e60 data points in time) to properly useBoxeJenkins approaches (i.e. AR, ARIMA and transfer func-tions models) (Box and Jenkins, 1976; Rothschild et al., 1996).

Predominantly flathead (Platycephalus spp.), mud crab(Scylla serrata), penaeid prawns (e.g., Penaeus esculentusand Penaeus plejebus), barramundi (Lates calcarifer) and mul-let (e.g. Mugil cephalus) were target species for the analyses.Where appropriate SOI and rainfall have been lagged as a com-mon method for such analyses (Sutcliffe, 1973; Drinkwateret al., 1991; Wilber, 1992) but this was done only when bio-logical meaningful (e.g. age of sexual maturity). Basic activi-ties of fish life cycles such as feeding and movement are mostlikely affected by environmental variability at shorter than an-nual or seasonal time scales. Therefore, monthly rainfall andmonthly catch data have also been included in the analyses.

Following a simple regression model the deviation of themean gross value of production (GVP) of Queensland’s totalestuary-dependent fish catch has been compared with averageannual coastal rainfall and El Nino events. This resulted in anestimation of potential loss caused by these events providinginformation for future model development.

4.4. CLIMPROD program

CLIMPROD (FAO, 1998) was used to see whether simplesurplus production models could adequately incorporate thelinks between environmental factors and fish catch. A fit assess-ment for the chosen models based on jackknife estimation ofthe parameters and of r2 (Efron and Gong, 1983) allowed a qual-ified test. Initial attempts were made to fit the Fox (1970),Schaefer (1957) and Pella and Tomlinson (1969) models usingcatch total effort for mullet (Mugil spp.), mud crab (Scylla ser-rata), barramundi (Lates calcarifer), king (Penaeus plejebus)and tiger prawns (Penaeus esculentus) and commercial estua-rine fish catch. The Schaeffer model is based on the Verhulstpopulation growth equation. Fox’s approach uses a logarithmicpopulation growth equation and Pella and Tomlinson’s ap-proach uses a generalized population growth equation.

Catch, effort and rainfall data were entered into the CLIM-PROD program and CLIMPROD’s expert system was used toselect an appropriate model based upon the known biology ofthe investigated species. This was done to see whether there isa simple surplus production model available to predict fish

496 J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

Fig. 3. The 8 different regions and their selected fish catch grids. Positive Pearson correlation values for monthly barramundi catch and rainfall (black, n ¼ 153) as

well as seasonal correlation values (grey, n ¼ 9) for the period 1988e2004 are shown for each region (*P < 0.05, **P < 0.01). Barramundi closure from Novem-

ber 1 (October in the Gulf) to February 1 has been excluded.

catch under environmental influence. Only simple modellingwas possible as there is a lack of ecological knowledge onthe investigated species (e.g., mortality rate, predators).CLIMPROD requires annual data series on catch and effortof a fishery on a single stock, and annual (or seasonal) data se-ries on an environmental variable. Standardization of effortwas not used in the CLIMPROD analyses, also this may im-prove the model (Evans et al., 1997). The appropriate modelis chosen, using a nonlinear regression routine and resultsare assessed with parametric and nonparametric tests. The co-efficient of determination (r2) was given by the program andthe jackknife method is used to give an indication of the ro-bustness of the model in which the percentage variation inr2 resulting from the elimination of a single data point fromthe set is calculated for all of the points (Efron and Gong,1983). The Food and Agricultural Organisation of the UnitedNations (FAO) suggests a conservative r2 value of above 90%for bivariate models and above 70% for multivariate models(FAO, 1998). The fit assessment is also based on residual anal-ysis and on the data set characteristics.

5. Results

5.1. El Nino events and estuarine fish catch

The preliminary analyses on total fish catch suggest that be-tween 20% and 30% of the catch variability might be

explained by El Nino events and the associated reduced rain-fall. Moreton Bay and Cairns total fish catch had r2 values of0.31 and 0.27, respectively, with yearly local rainfall recordsbetween 1988 and 2004. Comparison between penaeid prawnCPUE, rainfall and El Nino events indicated some furtherlinks. Three major El Nino periods can be identified for thelast 17 years showing significant correlation with Queenslandcoastal rainfall (Fig. 4). These periods are coincident with a re-duced CPUE for penaeid prawns (Family Penaeidae), mullet(e.g. Mugil cephalus), flathead (Platycephalus spp.) and the to-tal catch of estuary-dependent species, thus suggesting a linkbetween the SOI and some estuary-dependent species. Thesame holds for periods of very low average rainfall. However,analyses of the data showed strong influence of effort espe-cially on prawn catch. Correlation between average monthlyas well as annual temperature in Cairns and Moreton Bayand estuarine dependent fish catch were not significant.

The fluctuations in Queensland’s capture fisheries are coin-cident with El Nino and La Nina events and reduced rainfall. Ifthe loss is calculated from yearly Gross value of Production(GVP) during dry events against the mean GVP of the studyperiod, a total loss of A$ 65m during nine dry years versusa gain of A$ 61m in seven wet years was estimated. Thisseems reasonable as the fish catch explained about 66% ofthe GVP variability for the period 1988e2003. On a smallerscale the regional data for Cairns showed a similar patternwith an estimated loss during dry events relative to mean

497J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

GVP (1988e2004) of A$ 3.24m. The southeast Queenslandking (Penaeus plejebus), tiger (Penaeus esculentus, Penaeussemisulcatus) and bay prawn (Metapenaeus bennettae, Meta-penaeus insolitus) catches showed a strong reduction duringEl Nino events (1990, 1993, 1994, 1997) of up to 15 kgday�1 year�1 accounting for a total loss of A$ 70 000 during1988e2004 (Fig. 5).

5.2. Relationship between catch, rainfall and SOI

Separate correlation analyses were undertaken for flathead(Platycephalus spp.), mud crab (Scylla serrata), penaeidprawns (Family Penaeidae), barramundi (Lates calcarifer)and mullet (e.g. Mugil cephalus) to see differences of the cor-relation between the fish catch data and the interdependentvariables rainfall and SOI. Results obtained for annual barra-mundi catch and freshwater runoff showed positive but notsignificant correlation. There was also no significant correla-tion between yearly maximum or minimum temperature andcatch for any of the investigated species for the subregionCairns. Further analyses for temperature dependency havenot been undertaken.

The Moreton Bay and Cairns subregions both showed sig-nificant positive correlation between annual rainfall and totalfish catch (r ¼ 0.54 and 0.52; P < 0.05, n ¼ 17) (Fig. 6). Sig-nificant positive correlations also resulted from analysis be-tween mean annual coastal rainfall, SOI and total annualcommercial catches of mullet (Mugil spp.), barramundi (Lates

253035404550556065

1988

1990

1992

1994

1996

1998

2000

Years

C.P

.U

.E

. kg

/d

ay

-200-150-100-50050100150

SO

I

a

b

c

SOI

Fig. 4. SOI and southeast Queensland (a) tiger (Penaeus esculentus), (b) king

(Penaeus plejebus) and (c) bay prawn (Metapenaeus bennettae, M. insolitus)

catch.

0200400600800

1000120014001600

1988

1990

1992

1994

1996

1998

2000

2002

2004

years

rain

fall in

m

m

-200

-150

-100

-50

0

50

100

150

SO

I

SOI

rainfall

Fig. 5. Average coastal rainfall in Queensland and SOI over the time period

1988e2004 (r ¼ 0.62; P < 0.01, along the east coast and r ¼ 0.59;

P < 0.05, for the whole coast (n ¼ 17). El Nino events occurred in 1991/

1992, 1994/1995 and 1997/1998.

calcarifer), flathead (Platycephalus spp.) and mud crabs(Scylla serrata). Monthly rainfall and monthly barramundicatch were found to correlate significantly in all regions(Fig. 3) and in particular in the Herbert and Barron regions.Correlations for penaeid prawns, mud crabs and rainfallshifted from the south with significant positive to negative cor-relations in the north (Table 3). Furthermore, regression anal-yses have been performed for rainfall, SOI and selectedspecies groups, where the variation in SOI accounted for be-tween 41% and 49% of the variation (r2) in the catches of totalmullet (e.g. Mugil cepalus) and barramundi (L. calcarifer) inthe Gulf of Carpentaria whereas regional rainfall accountedfor between 55% and 81% of the variation (Fig. 7).

Long-term data for annual mullet catch (1945e1980) gaveno significant correlation with annual SOI (r ¼ 0.23;P > 0.05, n ¼ 35). However, the mullet CPUE from 1988e2004 showed a positive correlation with SOI (r ¼ 0.57;P < 0.05, n ¼ 17). Monthly mullet fish catch data for the timeperiod 1988 to 2004 were pooled into wet (NoveApr) anddry seasons (MayeOct) using annual coastal rainfall, seasonalrainfall for the Moreton South Coast region and seasonal SOIvalues. Mullet catch and CPUE correlated with SOI in MayeOct (range of r ¼ 0.61e0.65; P < 0.01; n ¼ 17). This correla-tion was weaker and non-significant for the wet season(r ¼ 0.36e0.39; P > 0.05, n ¼ 17) but still led to an overall sig-nificant correlation between annual SOI and catch (r ¼ 0.49e0.57; P < 0.05; n ¼ 17), which was contrary to the rainfalleCPUE relation. Therefore, the SOI (MayeOct) could be a usefulenvironmental parameter for mullet catch prediction.

5.3. CLIMPROD model

Analyses with CLIMPROD, using the Schaefer model gavepoor results with values of the goodness of fit parameter r2

varying from 0.02 to 0.20. When the SOI and rainfall datawere incorporated to landings and effort data, different modelswere obtained that had medium to high values r2 (0.30e0.87),and a wide range of jacknife r2 (0.001e0.85). Further statisti-cal analyses showed that there were small variations in the jac-knife r2 over the time series for barramundi (Lates calcarifer)east coast and mud crab (Scylla serrata) catches suggestingthat the model for these two species are robust. Followed bymullet (Mugil spp.) with a conventional r2 of 0.75 (Table 4).

1000

1500

2000

2500

3000

3500

4000

1988

1990

1992

1994

1996

1998

2000

2002

2004

years

catch

in

to

nn

es/year

20040060080010001200140016001800

rain

fall in

m

m

fish

rainfall

Fig. 6. Rainfall vs commercial fish catch (1988e2004) in Moreton Bay,

Queensland, Australia (r ¼ 0.54; P < 0.05, n ¼ 17).

498 J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

Table 3

Significant Pearson correlation coefficients (r) between commercial catches and rainfall in Queensland. Categories: n ¼ 13 annual (1988e2000), n ¼ 12 monthly

total, n ¼ 17 annual, n ¼ 204 (153 Barramundi open season) monthly for 1988e2004 (*P < 0.05; **P < 0.01). NP, North Peninsula; SP, South Peninsula; LC,

Lower Carpentaria; HNC, Herbert North Coast; ECC, East Central Coast; PSC, Port Curtis South Coast; MSC, Moreton South Coast

Species catch NP LC SP BNC HNC ECC PSC MSC SOI N

Barramundi East C. e e 0.64** 0.63** 0.77** e e e e 13

kg/day 0.68** 0.72** 0.71** 0.54* e e e e e 13

Gulf 0.85** 0.73** e e e e e e 0.77** 13

kg/day 0.89** 0.78** e e e e e e 0.57* 13

Annual 0.58* e e e e e e e 17

Seasonal e e e 0.58* 0.61* e e 0.84** e 12

kg/day e e e 0.67** e e 0.79** e 12

Monthly 0.16* 0.45** 0.42** 0.24** 0.21** 0.21** e 204

kg/day 0.27** 0.33** 0.25** 0.38** 0.39** 0.22** 0.17* e e 153

Mullet e e e e e e 0.54* 0.74** 0.63** 13

Annual e e e e e e e 0.78** 0.51** 17

kg/day e e e e e e e 0.67** 0.57* 17

Seasonal e e e e e e �0.61* e e 12

kg/day e e e e e e �0.62* e e 12

Monthly e e e e e e �0.18** �17** e 204

kg/day e e e e e e �0.15** �0.15** e 204

Flathead e e e e e 0.54* e 0.49* e 17

kg/day e e e e e e e 0.55* e 17

Mud Crab 0.74** 0.75** 0.79** 0.77** 0.72** e e e e 13

kg/day e e e 0.50 0.63** 0.55* e e 13

Annual 0.55* e 0.56* e e e e e e 17

kg/day 0.52* e e e 0.57* e e e e 17

Seasonal �0.74** e e e 0.69* 0.70** 0.86** e 12

kg/day e 0.72** 0.63* 0.63* 0.62* 0.80** 12

Monthly �0.24** �0.18** e e 0.19** 0.19** 0.22** 0.29** e 204

kg/day �0.18* 0.14* e 0.22** 0.26** 0.19** 0.34** 204

Tiger Prawn e e e 0.63* 0.51* e 0.53* e e 17

Seasonal e e e e e �0.58* e 0.63* e 12

Monthly �0.29** e �0.19** �0.16** e0.14* �0.27** 0.33** e 204

kg/day �0.38** e �0.45** e e e e 0.31** 204

Endeavour Prawn e e e 0.58* e e e e 0.61* 17

Seasonal �0.68* e �0.62* e e �0.69* 0.77** 0.89** e 12

kg/day �0.61* e �0.64* e e �0.63* 0.87* 0.91** e 12

Monthly �0.37** e �0.41** �0.22** e �0.21** 0.25** 0.38** e 204

kg/day �0.42** �0.39** �0.26** e �0.15** 0.28** 0.27** e 204

Greasy Prawn e e e e e e e 0.49* e 17

Seasonal e e e e e e 0.75** 0.91** e 12

kg/day e e e e e e e 0.91** e 12

Monthly e e e e e e 0.28** 0.35** e 204

kg/day e e e e e e 0.18** 0.49** e 204

6. Discussion and conclusion

6.1. Underlying mechanisms

Understanding the mechanisms underpinning relationshipsof climate parameters and fish production is essential to inter-pret the outcomes. Proposed mechanisms for the connectionbetween estuarine fishery species and rainfall include: (1) tro-phic linkages via changes to primary or secondary productionthat result from the addition of nutrients from terrestrial sour-ces (Aleem, 1972; Salen-Picard et al., 2002); (2) changes indistribution as a consequence of altered (expanded, reducedor connected) habitats (Loneragan and Bunn, 1999) and withit changes of catchability; and (3) changes in population

dynamics such as recruitment, growth, survival, abundance,assemblages and migration behaviour (Copeland, 1966; Pe-ters, 1982; Drinkwater and Frank, 1994; Gillanders and King-sford, 2002; Whitfield, 2005) as well as cohort or year-classstrength during the first year of life (Quinones and Montes,2001). The relationships between estuarine catch and rainfallare potentially confounded by other factors such as fishing ef-fort (Browder, 1985; da Silva, 1986). Their degree of influencedepends on the level of exploitation of the population by thefishery (Vance et al., 1985).

The analysis of the Queensland wild capture data between1988 and 2004 indicated that dry years, often associated withEl Nino, lead to reduced overall fish catch and wet years trans-lated to higher catches. This dependence may also reflect the

499J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

y = 0.6983x + 1.2019, r2 = 0.55y = 0.111x + 3.0663, r2 = 0.41

y = 0.1213x + 2.4402, r2 = 0.49y = 0.9086x - 0.0968, r2 = 0.81

2.4

2.6

2.8

3

3.2

3.4

2.8 2.9 3 3.1 3.2Log rainfall

Mullet

Barramundi

2.4

2.6

2.8

3

3.2

3.4

0 1 2 3

Mullet

Log SOI

Barramundi

Lo

g C

atch

Fig. 7. Relationship between (a) annual coastal rainfall and (b) SOI in Queensland, Australia and annual commercial catch between 1988 and 2000 of mullet (Mugilspp.), and barramundi (Lates calcarifer) from the Gulf of Carpentaria. Data for catch and flow are both on log10 axes. Lines are the least square regression lines.

known influence of SOI on some important non estuarine-de-pendent commercial fish species (e.g., mackerels (Yanez et al.,2002), tuna (Lehodey et al., 2003)). Similarly, on a smallerscale the regional data for southeast Queensland king, tigerand bay prawn catch showed a strong reduction during ElNino events for the study period. However, a detailed analysisof the prawn catch data demonstrated that regional differenceshave often contrasting effects on penaeid prawns: rainfall rela-tions reflected in positive correlation for South East Queens-land and negative correlations for North Queensland (Table3). Prawn species respond to rainfall with different discon-cordance in time, either positive or negative influencing theirontogenetic stages (Vance et al., 1998). Beside hydrologicaland biological differences, there are also differences in theprawn trawler fleet in Far North Queensland and the Gulf ofCarpentaria which mainly operates from April to May (Kailolaet al., 1993). The strong influence of effort on tiger and kingprawn catch makes a Queensland-wide simulation of climateeffects on these species difficult. On the other side, the positive

relationship between rainfall, SOI, catches, CPUE and CLIM-PROD outcomes for mud crabs (Scylla serrata), barramundi(Lates calcarifer) and mullet (e.g. Mugil cephalus) on a re-gional as well as state wide scale suggests that these environ-mental parameters may be taken for further modelling.Monthly and yearly average temperature did not have a signif-icant influence on the fish catch for two selected sites.

Regional differences in the rainfallecatch relationshipwere significant and reflected, for instance, the major fishingareas for mullet (e.g. Mugil cephalus) in the southeast, barra-mundi (Lates calcarifer) in the north, flathead (Platycephalusspp.) in the southeast and mud crab (Scylla serrata) on the eastcoast of Queensland and in the Gulf of Carpentaria. For in-stance, almost 80% of the Queensland mullet catch is har-vested from April to August (dry season) using about 40%of the mullet-harvesting effort. Monthly correlation betweenmullet and rainfall should therefore be negative (Table 3).The Moreton Bay region in the southeast of Queensland dom-inates harvest with about 70%, and the FrasereBurnett region

Table 4

Suggested production models by CLIMPROD to include rainfall or SOI. A good t-jackknife is given for all models. The MSY for mud crabs (Scylla serrata) may

be overestimated due to a constant increasing effort. CPUE is landing per unit effort kg/days/year; E, fishing effort; V, environmental variable; a, b, c and d are

constants. The MSY was estimated by the median of the MSY

Species Conventional r2 Jackknife r2 MSY in t (�SE) CPUE model Modification

Barramundi East coasta,c 0.93 0.85 246 (�40) aVb exp(EcVd) Age at recruitment 4 years

Barramundib,c 0.52 0.35 45 (�15) aV þ bE Age at recruitment 4 years

Mulleta,c 0.75 0.40 2076 (�101) a þ bV þ cV2 þ dE Age at recruitment 3 years

Mulleta,d 0.61 0.35 1909 (�135) (aV þ bV2)exp(cE ) Age at recruitment 3 years

Mullet May-Octb,d 0.34 0.09 1708 (�433) aV(1 þ bV)exp(cV(1 þ bV)E ) -

Mud Crabb,c 0.87 0.75 221 (�63) a þ bV þ cE Exploited year classes 3

Prawn Kingb,c 0.41 0.19 3176 (�281) (a þ bE )(1/c � 1) Exploited year classes 3

Prawn Tigerb,c 0.61 0.39 2464 (�313) (a þ bV)exp(cE ) Age at recruitment 1 year

Catch allb,c 0.65 0.47 na aV þ bV2 þ c Exploited year classes 3

a 1988e2000.b 1988e2004.c Rainfall.d SOI.

500 J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

provides about 25% of the harvest (Williams, 2002). This isreflected by a significant relationship between mullet catchand annual rainfall in the Moreton Bay region (Table 3).The general reproductive pattern of M. cephalus involves mi-gration from either fresh or estuarine waters to offshore waterswhere they spawn in large schools. Larvae and fry then mi-grate to inshore estuaries where they inhabit shallow, warmwater in the intertidal zone. This reproductive cycle suggeststhat wetter periods between May and August may stimulatethe migration of mullet (Mugil spp.) out of the estuaries andtherefore increases their catchability, since most of the mulletis caught along beaches and not in estuaries. Dryer butwarmer wet seasons would, however, increase algal productiv-ity in the estuaries and therefore strengthen the 0þ year-classwith increased catches lagged by the period that it takes indi-viduals of the species to ‘recruit’ to the fishery, demonstratinga lagged response. A similar mechanism operates for the in-creased catches of mud crabs (S. serrata), which are notcaught in the fishery until they are at least 12e24 monthsold because of legal size limit regulations. Higher rainfalland hence river flow stimulates the downstream movementof mud crabs (Hill et al., 1982) and this could increase theircatchability in the lower estuary and bay. The reduction innumbers of subadult and adult crabs in the river systemsmay also enhance the survival of juveniles because of reducedcannibalism and competition for burrows which may be theexplanation for the strong correlation between 1-year laggedrainfall and mud crab (S. serrata) catch. Such a lagged effectof river flow on catches has also been recorded for the rela-tionship between annual rainfall and barramundi catches inthe Gulf of Carpentaria and some estuaries on the east coast,due to enhanced productivity and increased survival and/orgrowth of the juvenile stages (Staunton-Smith et al., 2004).The lagged response of fish species to rainfall changes is animportant characteristic for a number of species. Overall,there can be positive effects of rainfall (availability of food)or negative effects (higher mortality of juveniles, reductionof usable nursery habitat) because of changes in salinityand/or turbidity (Staples, 1980; Potter et al., 1991) dependingon genetic adaptation. These arguments collectively suggestlinked mechanisms between productivity, rainfall and catch-ability (Kimmerer, 2002).

However, to prove underlying environment-recruitmentcorrelations is extremely difficult. There is as yet insufficientknowledge about impacts of climate changes on regionalocean currents and about physical-biological linkages. Theargument for including climate factors in assessments isclearly strengthened if the mechanistic links are better un-derstood (see also Fig. 2). We expect rainfall and tempera-ture to alter fish species composition. Further analyses inthis direction are necessary. Analysis of relationships be-tween catch and environmental variables are often used inobservational studies to derive insights into the factorsdriving the distribution and abundance of fisheries speciesat the whole estuary scale (Tyler, 1992) and provide direc-tions for research and modelling approaches (Planqueet al., 2004).

6.2. Modelling

Simple modelling of economic impacts of climate changecan be estimated by running a regression of SOI and rainfallwith fish catch over time (Xie and Hsieh, 1989; Beamishand Bouillon, 1993; Aaheim and Sygna, 2000). This methodhas been applied successfully to a number of studies (Vanceet al., 1985; Loneragan and Bunn, 1999; Growns and James,2005) and can provide first ideas on the relation between theseparameters and fish catch. However, the regression approachhas been criticised because of: (1) the confounding effects ofstock size and fishing pressure (Walters and Collie, 1988);(2) the likely non-linearity of linking mechanisms (Baumann,1998) and the probability of multiple mechanisms; (3) the lackof ability to prove causality (Quinones and Montes, 2001); and(4) their uncertain predictive capability as a consequence oflong-term climatic variation or human-induced changes (e.g.,habitat loss, pollution).

The results from the CLIMPROD program were overall en-couraging and will allow incorporation, for example, into bio-economic models. The application of the CLIMPRODprogram to Queensland yield and effort and seasonal rainfallinputs resulted in validated modelling of the fishery forsome estuary-dependent species and the analyses can assistin refining hypotheses developed from life history assessment,and provide guidance as to where further research should befocused. The case studies suggested that regional and seasonaldifferences in rainfall and landings have to be considered butthat the positive relationship of certain fish species with rain-fall along the coast is valid on a broad scale. However, themodels including rainfall or SOI as environmental variables(Aksnes et al., 1995) also showed that more refined modellingtools are required to provide better outcomes. Simple surplusproduction models have been criticized because they lack bi-ological realism, resulting in poor predictive ability. One ofthe weaknesses of the biomass production modelling approachis that changes in community composition will not be de-tected. For example, they do not consider the structure ofthe stock by age, size or by species. In many cases, however,more sophisticated age-structured models do not perform bet-ter, owing to difficulties in additional parameters estimation(Ludwig and Walters, 1985). Dynamic software can achievevalid outcomes for the efficient management of a fishery asthey include several dynamic (non-equilibrium) models (FAO,1996; FAO, 1998). The available fish catch data set and infor-mation on species biology limits the use of other models.

Some further problems when analysing the data include:(1) the failure to account for autocorrelation in the time-series;(2) over-fitting of a function due to unknown behaviour outsideof the data sample; and (3) the identification of spurious corre-lations where the variables are statistically related but are notcausally linked (Hilborn and Walters, 1992). A number of statis-tical techniques are available to deal with autocorrelation al-though they tend to reduce the test power (Drinkwater, 1987).The risks of over-fitting can be reduced by careful exploratorycorrelation analyses and applying corrections to threshold prob-abilities for multiple comparisons (Carscadden et al., 2000).

501J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

The choice of the environmental variable is often the key factorto avoid spurious correlation. In general, even in the case of sur-plus production models, a minimum knowledge of the stock andof the species biology is required. Further analysis of the datamay be via low and high frequency analysis (e.g. ARIMA)(Pyper and Peterman, 1998). ARIMA (autoregressive integratedmoving average) models can remove the autocorrelation and/ormove average structure of the series transforming it into whitenoise. White noise can be defined as a stochastic process bya marginal distribution function. Transfer functions allow tobe associated with environmental or other external variables,or intervention analysis may help to detect anomalous events(Rothschild et al., 1996). However, consistent long-term datasets for seasonally-based analysis with more than 50 data pointsfor effective time-series analysis are currently not available forQueensland.

A more practical modelling approach alongside the use ofother statistical tools (e.g., CLIMLAB 2000, GAM andGLM) has been shown by Kasulo and Perrings (2004) whostudied the influence of changes in environmental conditionsfor Lake Malawi. They suggested a modified GordoneSchae-fer model (Gordon, 1954; Schaefer, 1957) to include both en-vironmental and bio-economic diversity variables (Simonitand Perrings, 2005). Since economics are an important aspectdriving fishing activities, bio-economic modelling is muchmore realistic than purely biological (or purely economic) ap-proaches (FAO, 1993, 1997; King, 1995; Helstad, 2000; Ulrichet al., 2001). However, a single model cannot describe all as-pects of the environment that are relevant. A coupled model, inwhich different types of models are used to explain differentparts of the system, can provide better outcomes. One impor-tant part of this approach is a trophic model (Pauly et al.,1996) summarises the species present in an ecosystem, theirabundances, and the quantities of one species consumed by an-other. For example, estuarine fish such as barramundi (Latescalcarifer), king threadfin (Polydactylus sheridani) and catfish(Ariidae) are major predators of juvenile tiger prawns (Pe-naeus esculentus) (Salini et al., 1990). Such a model should in-clude movements, feeding and survival rate. The integratedenergy balance ECOPATH with ECOSIM software (Walterset al., 1997) for dynamic simulation modelling can determinethe long-term catch rates and biomass that result under varyingdegrees of fishing mortality (Christensen and Walters, 2004;Pitcher, 2005). ECOSPACE, the spatial version of ECOSIMcan also dynamically allocate biomass across a grid map (Wal-ters et al., 1997; Okey and Pauly, 1999). However, the lack oflocal information for the modelling would create an importantuncertainty for the simulations with ECOPATH and ECOSIMfor our study. Better fitting can be achieved with time seriesdata on biomass, fishing mortality and survey informationwhen available.

Our case study identified a clear link between estuarine fishcatch and climate variables on a broad scale but also showedthat there are individual differences between species and re-gions to be considered in modelling approaches. In conclusion,the fisheries should not be managed without including proac-tive responses to changes in rainfall and the SOI and their

predicted future trends. The SOI can be predicted with someaccuracy for up to six months including some estimation forrainfall. Such prediction could be a significant element in anenvironmentally-sensitive management policy. It is necessaryto identify a more comprehensive management policy thatwill ensure sustainability even under conditions of higher en-vironmental pressure, e.g., from more frequent and more ex-treme El Nino conditions resulting in a reduction of rainfall.Predictions of SOI and rainfall are possible and should notonly be used in agricultural forecasts. Overexploitation andeconomic losses can be avoided when planning ahead by tar-geting fish species which are less influenced by El Nino eventsthan others, e.g., mullet (Mugil spp.) versus whiting (Sillagospp.). The multi-species fishery in Queensland would havethe opportunity to undergo such shifts to increase both stablefish stocks and economic outcome.

Acknowledgments

We thank the Queensland Department of Primary Indus-tries, Assessment and Management Unit for providing thedata from the CFISH database, the Australian Bureau of Me-teorology for climate data and the Queensland Department ofNatural Resources for providing data on flows in the BarronRiver. Thanks to Pierre Freon, Jerrold G. Norton, LeonardOlyott, Ann Farell and Verena Schrameyer for commentsand helpful discussion.

References

Aaheim, A., Sygna, L., 2000. Economic Impacts of Climatic Change on Tuna

Fisheries in Fiji Islands and Kiribati. Centre for International Climate and

Environmental Research, Blinder, Norway, 21 pp.

Abbs, D. (Ed.), 2002. Climate Change and Australia’s Coastal Communities.

CSIRO, Atmospheric, Aspendale, 8 pp.

Aksnes, D.L., Ulvestad, K.B., Balino, B.M., Berntsen, J., Egge, J.K.,

Svendsen, E., 1995. Ecological modelling in coastal waters: towards

predictive physical-chemical-biological simulation models. Ophelia 41,

5e36.

Alber, M., 2002. A conceptual model of estuarine freshwater inflow manage-

ment. Estuaries 25, 1246e1261.

Aleem, A.A., 1972. Effect of river outflow management on marine life. Marine

Biology 15, 200e208.

Baumann, M., 1998. The fallacy of the missing middle: physics / . / fish-

eries. Fisheries Oceanography 7, 63e65.

Beamish, R.J., Bouillon, D.R., 1993. Pacific salmon production trends in rela-

tion to climate. Canadian Journal of Fisheries and Aquatic Sciences 50,

1002e1016.

Blaber, S.J.M., Blaber, T.G., 1980. Factors affecting the distribution of juve-

nile estuarine and inshore fish. Journal of Fish Biology 17, 143e162.

Box, G.E.P., Jenkins, G.M., 1976. Time Series Analysis, Forecasting and Con-

trol. Holden-Day, Inc., San Francisco, 575 pp.

Browder, J.A., 1985. Relationship between pink shrimp production on the Tor-

tugas grounds and water flow patterns in the Florida Everglades. Bulletin

of Marine Science 37, 839e856.

Browder, J.A., Zein-Eldin, Z.P., Crailes, M., Robblee, M.B., Wong, S.,

Jackson, T.L., Johnson, D., 2002. Dynamics of pink shrimp (Farfantepe-naeus duorarum) recruitment potential in relation to salinity and tempera-

ture in Florida Bay. Estuaries 25, 1355e1371.

Byrne, C.J., Poole, R., Dilliane, M., Rogan, G., Whelan, K.F., 2002. Temperoral

and environmental influences on the variation in sea trout (Salmo trutta L.)

502 J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

Smolt migration in the Burrishoole system in the west of Ireland from 1971 to

2000. Fisheries Research 66, 85e94.

Carscadden, J.E., Frank, K.T., Leggett, W.C., 2000. Evaluation of an environ-

ment-recruitment model for capelin (Mallotus villosus). ICES Journal of

Marine Science 57, 412e418.

Chen, Y.L., Chen, H.Y., Tzeng, W.N., 1994. Reappraisal of the importance of

rainfall in affecting catches of Anguilla japonica elvers in Taiwan. Austra-

lian Journal of Marine Freshwater Research 45, 185e190.

Chen, W.Z., Zheng, Y.Z., Chen, Y.Q., Mathews, C.P., 1997. An Assessment of

Fishery Yields from the East China Sea Ecosystem. Marine Fisheries Re-

view 59, 1e7.

Christensen, V., Walters, C.J., 2004. Ecopath with Ecosim: methods, capabil-

ities and limitations. Ecological Modelling 172, 109e139.

Copeland, B.J., 1966. Effects of decreased river flow on estuarine ecology.

Journal of Water Pollution Control Federation 38, 1831e1839.

Currie, D.R., Small, K.J., 2005. Macrobenthic community responses to long-

term environmental change in an east Australian sub-tropical estuary. Es-

tuarine, Coastal and Shelf Science 63, 315e331.

da Silva, J., 1986. River runoff and shrimp abundance in a tropical coastal eco-

system e the example of the Sofala bank (central Mozambique). In:

Skreslet, S. (Ed.), NATO Advanced Research Workshop on the Role of

Freshwater Outflow in Coastal Marine Ecosystems. Springer-Verlag,

Berlin, pp. 329e344.

Dent, S. (Ed.), 2002. Queensland Fisheries Profile. Queensland Department of

Primary Industries and Fisheries, Brisbane, 45 pp.

Drinkwater, K.F., 1987. Sutcliffe revisited. Previously published correlations

between fish stocks and environmental indices and their recent perfor-

mance. In: Perry, I.R.I., Frank, K.T. (Eds.), Environmental Effects Recruit-

ment to Canadian Fish Stocks, 1556. Canadian Technical Report on

Fisheries Aquatic Science, pp. 41e61.

Drinkwater, K.F., Frank, K.T., 1994. Effects of river regulation and diversion

on marine fish and invertebrates. Aquatic Conservation: Freshwater and

Marine Ecosystems 4, 135e151.

Drinkwater, K.F., Harding, G.C., Vass, W.P., Gauthier, D., 1991. The relation-

ship of Quebec lobster landings to freshwater runoff and wind storms. In:

Therriault, J.C. (Ed.), The Gulf of St. Lawrence: Small Ocean or Big

Estuary? Canadian Special Publication of Fisheries and Aquatic Sciences,

113, pp. 179e187.

Efron, B., Gong, G., 1983. A leisurely look at the bootstrap, the jackknife, and

cross-validation. The American Statistician 37, 36e48.

Eide, A., Heen, K., 2002. Economic impacts of global warming, a study to the

fishing industry in North Norway. Fisheries Research 56, 261e274.

Evans, C.R., Lockwood, A.P.M., Evans, A.J., 1995. Associations between sea

temperature, catch per unit of fishing effort, and yield in the Bermuda

spiny lobster fishery, 1975e89. Marine and Freshwater Research 46,

809e818.

Evans, C.R., Opnai, L.J., Kare, B.D., 1997. Fishery ecology and oceanography

of the prawn Penaeus merguiensis (de Man) in the Gulf of Papua: estima-

tion of maximum sustainable yield and modelling of yield, effort and rain-

fall. Marine and Freshwater Research 48, 219e228.

FAO, 1993. BEAM4 Analytical Bio-economic Simulation of Spacestructured

Multispecies and Multifleet Fisheries. In: Description of the Model. User’s

manual, vol. 1. Food and Agriculture Organization of the United Nations,

Rome, pp. 46e186.

FAO, 1994. SPATIAL Space-time Dynamics in Marine Fisheries. A Software

Package for Sedentary Species. User’s Manual. Food and Agriculture Or-

ganization of the United Nations, Rome, 116 pp.

FAO, 1997. Bioeconomıa Pesquera. Teoria, Modelacion y Manejo. Food and

Agriculture Organization of the United Nations, Rome, 176 pp.

FAO, 1996. Biomass Dynamic Models. User’s Manual. FAO Computerized In-

formation Series (Fisheries). Food and Agriculture Organization of the

United Nations, Rome, 62 pp.

FAO, 1998. CLIMPROD (version 2): Experimental Interactive Software for

Choosing and Fitting Surplus Production Models Including Environmental

Variables. Computerized Information Series (Fisheries). Food and Agricul-

ture Organization of the United Nations, Rome, 121 pp.

FAO, 2000. The Significance of Environmental Information in the Manage-

ment of Oil Sardine Fishery in the Bali Strait. Fishery and Management

of Bali Sardinella (Sardinella lemuru) in Bali Strait. Food and Agriculture

Organization of the United Nations, Rome, 43e52.

Forbes, A.T., Cyrus, D.P., 1993. Biological effects of salinity gradient reversals

in a southeast African estuarine lake. Netherlands Journal of Aquatic Ecol-

ogy 27, 265e272.

Galindo-Bect, M.S., Glenn, E.P., Page, H.M., Fitzsimmons, K., Galindo-

Bect, L.A., Hernandez-Ayon, J.M., Petty, R.L., Garcia-Hernandez, J.,

Moore, D., 2000. Penaeid shrimp landings in the upper gulf of California

in relation to Colorado River freshwater discharge. Fishery Bulletin 98,

222e225.

Gammelsrod, T., 1992. Variation in shrimp abundance on the Sofala Bank,

Mozambique and its relation to the Zambezi River runoff. Estuarine,

Coastal and Shelf Science 35, 91e103.

Geoscience Australia, 2005. Maps of Australia. 2005. Commonwealth of Aus-

tralia, Canberra.

Gillanders, B.M., Kingsford, M.J., 2002. Impact of changes in flow of fresh-

water on estuarine and open coastal habitats and the associated organisms.

Oceanography and Marine Biology: An Annual Review 40, 233e309.

Glaister, J.P., 1978. The impact of river discharge on distribution and produc-

tion of the school prawn Metapenaeus macleayi (Haswell) (Crustacea: Pe-

naeidae) in the Clarence River region, northern New South Wales.

Australian Journal of Marine and Freshwater Research 28, 311e323.

Gopurenko, D., Hughes, J.M., 2002. Regional patterns of genetic structure

among Australian populations of the mud crab, Scylla serrata (Crustacea:

Decapoda): evidence from mitochondrial DNA. Marine and Freshwater

Research 53, 849e857.

Gordon, H.S., 1954. The economic theory of a common property resource: the

fishery. Journal of Political Economy 62, 124e142.

Growns, I., James, M., 2005. Relationships between river flows and recrea-

tional catches of Australian bass. Journal of Fish Biology 66, 404.

Gunter, G., Hildebrand, H.H., 1954. The relation of total rainfall of the state

and catch of the marine shrimp (Penaeus setiferus) in Texas waters. Bulle-

tin of Marine Science of the Gulf and Caribbean 4, 95e103.

Hannesson, R., 1993. Bioeconomic Analysis of Fisheries. Halsted Press, New

York, 138 pp.

Hart, D.D., Finelli, C.M., 1999. Physical-biological coupling in streams: the

pervasive effects of flow on benthic organisms. Annual Review of Ecology

and Systematics 30, 363e395.

Helstad, K., 2000. Management of Herring, Capelin and Cod in the Greater

Barents Sea e Economic Optimal Management from a Norwegian Point

of View. The Norwegian College of Fishery Science. University of

Tromsø, Tromsø, 22 pp.

Hilborn, R., Walters, C.J., 1992. Quantitative Fisheries Stock Assessment:

Choice Dynamics and Uncertainty. Chapman and Hall, New York, 570 pp.

Hill, B.J., Williams, M.J., Dutton, P., 1982. Distribution of juvenile, subadult

and adult Scylla serrata (Crustacea: Portunidae) on tidal flats in Australia.

Marine Biology 69, 117e120.

Hill, K., 2004. Stripped Mullet (Mugil cephalus). Smithsonian Marine Station,

Fort Pierce, 5 pp.

Hughes, L., 2003. Climate change and Australia: trends, projections and im-

pacts. Austral Ecology 28, 423e443.

IPCC, 2001. Summary for policymakers: climate change 2001: impacts, adap-

tation and vulnerability. In: McCarthy, J.J., Cauziani, O.F., Leary, N.A.,

Dokken, D.J., White, K.S. (Eds.), Climate Change 2001: Impacts, Adap-

tion, and Vulnerability: Contribution of Working Group II to the Third As-

sessment Report of the Intergovernmental Panel on Climate Change.

Intergovernmental Panel on Climate Change. Cambridge University Press,

Cambridge, 913 pp.

Kailola, P.J., Williams, M.J., Stewart, P.C., Reichelt, R.E., McNee, A.,

Grieve, C., 1993. Australian Fisheries Resources. Commonwealth of Aus-

tralia Bureau of Resource Sciences and the Fisheries Research & Develop-

ment Corporation, Canberra, 422 pp.

Kasulo, V., Perrings, C., 2004. Fishing Down the Value Chain: Modelling the

Impact of Biodiversity Loss in Freshwater Fisheries e The Case of

Malawi. Arizona State University, Tempe, 23 pp.

Kimmerer, W.J., 2002. Effects of freshwater flow on abundance of estuarine

organisms: physical effects or trophic linkages? Marine Ecology Progress

Series 243, 39e55.

503J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

King, M., 1995. Fisheries biology. Assessment and management. Fishing

News Books, 341 pp.

Klyashtorin, L.B., 1998. Long term climate change and main commercial

fish production in the Atlantic and Pacific. Fisheries Research 37,

115e125.

Lea, R., 2000. Observations on fishes associated with the 1997e1998 El Nino

off California. CalCOFI Report 41, 117e129.

Lehodey, P., Chai, F., Hampton, J., 2003. Modelling climate-related variability

of tuna populations from a coupled ocean-biogeochemical- populations dy-

namics model. Fisheries Oceanography 12, 483e494.

Lloret, J., Palomera, I., Salat, J., Sole, I., 2004. Impact of freshwater input and

wind on landings of anchovy (Engraulis encrasicolus) and sardine (Sar-dina pilchardus) in shelf waters surrounding the Ebre (Ebro) River delta

(north-western Mediterranean). Ficheries Oceanography 13, 102e110.

Loneragan, N.R., Bunn, S.E., 1999. River flows and estuarine ecosystems:

implications for coastal fisheries from a review and a case study of

the Logan River, southeast Queensland. Australian Journal of Ecology

24, 431e440.

Ludwig, A., Walters, C.J., 1985. Are age-structured models appropriate for

catch-effort data? Canadian Journal of Fisheries and Aquatic Sciences

42, 1066e1072.

Norton, J.G., McLain, D.R., 1994. Diagnostic patterns of seasonal and interan-

nual temperature variation off the west coast of the United States: local and

remote large-scale atmospheric forcing. Journal of Geophysical Research

99, 16019e16030.

Okey, T., Pauly, D. (Eds.), 1999. A Trophic Mass-Balance Models of Alaska’s

Prince William Sound Ecosystem for the Post-Spill Period 1994e1996,

second ed. Fisheries Center Research Report, The University of British

Columbia, Vancouver, pp. 137.

Pauly, D., Christensen, V., Haggan, N., 1996. Mass-Balance Models of North-

eastern Pacific Ecosystems. Fisheries Centre. University of British Colum-

bia, Vancouver, 131 pp.

Peters, J.C., 1982. Effects of river and streamflow alteration on fishery re-

sources. Fisheries Research 7, 20e22.

Pitcher, T.J., 2005. Spatial Simulations with Ecospace. In: Haggan, N.,

Pitcher, T.J. (Eds.), Ecosystem Simulation Models of Scotland’s West Coast

and Sea Lochs. UBC Fisheries Centre Research Reports, 13, pp. 47e53.

Potter, I.C., Manning, R.J.G., Loneragan, N.R., 1991. Size, movements, distri-

bution and gonadal stage of the western king prawn (Penaeus latisulcatus)

in a temperate estuary and local marine waters. Journal of Zoology (Lon-

don) 223, 419e445.

Powell, G.L., Matsumoto, J., Brock, D.A., 2002. Methods for determining

minimum freshwater inflow needs of Texas Bays and estuaries. Estuaries

25, 1262e1274.

Power, S., Tseitkin, F., Meht, V., Lavery, B., Torok, s., Holbrook, N., 1999. De-

cadal climate variability in Australia during the twentieth century. Interna-

tional Journal of Climatology 19, 169e184.

Pyper, B.J., Peterman, R.M., 1998. Comparison of methods to account for au-

tocorrelation in correlation analyses of fish data. Canadian Journal of Fish-

eries and Aquatic Sciences 55, 2127e2140.

Quinn, R.H., 1987. Analysis of fisheries logbook information from the Gulf of

Carpentaria, Australia. In: Copeland, J.W., Grey, D.L. (Eds.), Management

of Wild and Cultured Sea Bass/Barramundi (Lates calcarifer). Australian

Centre for International Agricultural Research, Canberra, pp. 92e95.

Quinones, R.A., Montes, R.M., 2001. Relationship between freshwater input

to the coastal zone and the historical landing of the benthic/demersal

fish Eleginops maclovinus in central-south Chile. Fisheries Oceanography

10, 311e328.

Robertson, A.I., Duke, N.C., 1990. Recruitment, growth and residence time of

fishes in a tropical Australian mangrove system. Estuarine, Coastal and

Shelf Science 31, 723e743.

Robins, J.B., Halliday, I.A., Staunton-Smith, J., Mayer, D.G., Sellin, M.J.,

2005. Freshwater-flow requirements of estuarine fisheries in tropical Aus-

tralia: a review of the state of knowledge and application of a suggested

approach. Marine and Freshwater Research 56, 343e360.

Roessig, J., Woodley, C., Cech, J., Hansen, L., 2004. Effects of global climate

change on marine and estuarine fishes and fisheries. Review in Fish Biol-

ogy and Fisheries 14, 251e275.

Rothschild, B.J., Smith, S.G., Li, H., 1996. The application of time series anal-

ysis to fisheries populations assessment and modeling. In: Gallucci, V.F.,

Saila, S.B., Gustafson, D.J., Rothschild, B.J. (Eds.), Stock Assessment:

Quantitative Methods and Applications for Small-Scale Fisheries. CRC

Press, Florida, pp. 354e402.

Ruello, N., 1973. The influence of rainfall on the distribution and abundance of

the school prawn in the Hunter River region (Australia). Marine Biology

23, 221e228.

Salen-Picard, C., Darnaude, A.M., Arlhac, A.M., Harmelin-Viven, M.L., 2002.

Fluctuations of macrobenthic populations: a link between climate-driven

river run-off and sole fishery yields in the Gulf of Lions. Oecologia 133,

380e388.

Salini, J.P., Blaber, S.J.M., Brewer, D.T., 1990. Diets of piscivorous fishes in

a tropical estuary, with special reference to predation on penaeid prawns.

Marine Biology 105, 363e374.

Schaefer, M.B., 1957. A study of the dynamics of the fishery for yellow tuna in

the eastern tropical pacific ocean. Bulletin of the Inter-American Tropical

Tuna Commission 2, 247e268.

Schwartzlose, R.A., Alheit, J., Bakun, A., Baumgartner, T.R., Cloete, R.,

Crawford, R.J.M., Fletcher, W.J., Green-Ruiz, Y., Hagen, E.,

Kawasaki, T., Lluch-Belda, D., Lluch-Cota, S.E., MacCall, A.D.,

Matsuura, Y., Nevarez-Martinez, M.O., Parrish, R.H., Roy, C., Serra, R.,

Shust, K.V., Ward, M.N., Zuzunaga, J.Z., 1999. Worldwide large-scale

fluctuations of sardine and anchovy populations. South African Journal

of Marine Science 21, 289e347.

Sklar, F.H., Browder, J., 1998. Coastal environmental impacts brought about

by alterations to freshwater flow in the Gulf of Mexico. Environmental

Management 22, 547e562.

Simonit, S., Perrings, C., 2005. Indirect economic indicators in bio-economic

fishery models: agricultural price indicators and fish stocks in Lake Victo-

ria. ICES Journal of Marine Science 62, 483e492.

Smith, L.D., 1990. Patterns of limb loss in the blue crab, Callinectes sapidus-

Rathbun, and the effects of autotomy on growth. Bulletin of Marine Sci-

ence 46, 23e36.

Sparre, P., Venema, S.C., 1997. Introduc~ao a avaliac~ao de mananciais de

peixes tropicais. In. FAO Documentas Tecnologica Pescas, 306/1 Rev 2.

(Parte 1 and 2), pp. 404e494.

Staples, D.J., 1980. Ecology of juvenile and adolescent banana prawns,

Penaeus merguiensis, in a mangrove estuary and adjacent off-shore areas

of the Gulf of Carpentaria I. Immigration and settlement of postlarvae.

Australian Journal of Marine and Freshwater Research 31, 635e652.

Staples, D.J., Vance, D.J., 1986. Emigration of juvenile banana prawns Pe-naeus merguiensis from a mangrove estuary and recruitment to offshore

areas in the wet-dry tropics of the Gulf of Carpentaria, Australia. Marine

Ecology Progress Series 27, 239e252.

Staunton-Smith, J., Robins, J.B., Mayer, D.G., Sellin, M.J., Halliday, I.A.,

2004. Does the quantity and timing of fresh water flowing into a dry trop-

ical estuary affect year-class strength of barramundi (Lates calcarifer)?

Marine and Freshwater Research 55, 787e797.

Sutcliffe, W.H.J., 1973. Correlations between seasonal river discharge and lo-

cal landings of American lobster (Homarus americanus) and Atlantic hal-

ibut (Hippoglossus hippoglossus) in the Gulf of St. Lawrence. Journal of

the Fisheries Research Board of Canada 30, 856e859.

Tyler, A.V., 1992. a context for recruitment correlation: why marine fisheries

biologists should still look for them. Fisheries Oceanography 1, 97e107.

Ulrich, C.D., Gascuel, M.R., Dunn, B., Le Gallic, B., Dintheer, C., 2001. Es-

timation of technical interactions due to the competition for resource in

a mixed-species fishery, and the typology of fleets and metiers in the En-

glish Channel. Aquatic Living Resources 14, 267e281.

Vance, D.J., Haywood, M.D.E., Heales, D.S., Kenyon, R.A., Loneragan, N.R.,

1998. Seasonal and annual variation in abundance of postlarval and juve-

nile banana prawns Penaeus merguiensis and environmental variation in

two estuaries in tropical northeastern Australia: a six year study. Marine

Ecology Progress Series 163, 21e36.

Vance, D.J., Staples, D.J., Kerr, J.D., 1985. Factors affecting year-to-year var-

iation in the catch of banana prawns (Penaeus merguiensis) in the Gulf of

Carpentaria, Australia. Journal Conseil International pour l’Exploration de

la Mer 42, 83e97.

504 J.-O. Meynecke et al. / Estuarine, Coastal and Shelf Science 69 (2006) 491e504

Walsh, K., Hennessy, K., Jones, R., Pittock, A.B., Rotstayn, L., Suppiah, R.,

Whetton, P., 2000. Climate Change in Queensland Under Enhanced Green-

house Conditions, Second Annual Report 1998e1999. Aspendale, CSIRO

Atmospheric Research, 129 pp.

Walters, C.J., Christensen, V., Pauly, D., 1997. Structuring dynamic model of

exploited ecosystems from trophic mass-balance assessments. Reviews in

Fish Biology and Fisheries 7, 139e192.

Walters, C.J., Collie, J.S., 1988. Is research on environmental factors useful to

fisheries management? Canadian Journal of Fisheries and Aquatic Sci-

ences 45, 1848e1854.

Watts, R.J., Johnson, M.S., 2004. Estuaries, lagoons and enclosed embay-

ments: habitats that enhance population subdivision of inshore fishes. Ma-

rine and Freshwater Research 55, 641e651.

Whitfield, A.K., 1994. Abundance of larval and 05 juvenile marine

fishes in the lower reaches of three southern African estuaries with

differing freshwater inputs. Marine Ecology Progress Series 105,

257e267.

Whitfield, A.K., Harrison, T.D., 2003. River flow and fish abundance in a South

African estuary. Journal of Fish Biology 62, 1467e1472.

Whitfield, A.K., 2005. Fishes and freshwater in southern African estuaries e

a review. Aquatic Living Resources 18, 275e289.

Wilber, D., 1992. Associations between freshwater in inflows and oyster pro-

ductivity in Apalachicola Bay, Florida. Estuarine, Coastal and Shelf Sci-

ence 35, 179e190.

Williams, L.E. (Ed.), 2002. Queensland’s Fisheries Resources: Current Condi-

tion and Recent Trends 1988e2000. Queensland Department of Primary

Industries and Fisheries, Brisbane, 180 pp.

Xie, L., Hsieh, W.W., 1989. Predicting the return migration routes of the

Fraser River sockeye salmon (Oncorhynchus nerka). Canadian Journal of

Fisheries and Aquatic Sciences 46, 1287e1292.

Yanez, E., Silva, C., Nieto, K., Barbieri, M.A., Martinez, G., Ramırez, B.,

2002. Prediction of Probable Fishing Grounds in Northern Chile from Pe-

lagic Fisheries Distributions and Environmental Conditions. Global Ocean

Ecosystem Dynamics Report 16, Plymouth, 24e26.