Reprint of: The ecophysiology of Sprattus sprattus in the Baltic and North Seas

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

The ecophysiology of Sprattus sprattus in the Baltic and North Seas

Myron A. Peck a,⇑, Hannes Baumann a,1, Matthias Bernreuther a,2, Catriona Clemmesen b,Jens-Peter Herrmann a, Holger Haslob b, Bastian Huwer c, Philipp Kanstinger a, Fritz W. Köster c,Christoph Petereit b, Axel Temming a, Rudi Voss b,3

a Center for Earth System Research and Sustainability, Institute for Hydrobiology and Fisheries Science, University of Hamburg, Olbersweg 24, 22767 Hamburg, Germanyb IFM-GEOMAR, Leibniz-Institute for Marine Research at the University of Kiel, Düsternbrooker Weg 20, 24105 Kiel, Germanyc DTU Aqua, National Institute of Aquatic Resources, Technical University of Denmark, Charlottenlund Slot, Jægersborg Allé 1, 2920 Charlottenlund, Denmark

a r t i c l e i n f o

Article history:Received 1 October 2010Received in revised form 10 January 2012Accepted 16 April 2012Available online 25 April 2012

a b s t r a c t

The European sprat (Sprattus sprattus) was a main target species of the German GLOBEC program thatinvestigated the trophodynamic structure and function of the Baltic and North Seas under the influenceof physical forcing. This review summarizes literature on the ecophysiology of sprat with an emphasis ondescribing how environmental factors influence the life-history strategy of this small pelagic fish. Onto-genetic changes in feeding and growth, and the impacts of abiotic and biotic factors on vital rates are dis-cussed with particular emphasis on the role of temperature as a constraint to life-history scheduling ofthis species in the Baltic Sea. A combination of field and laboratory data suggests that optimal thermalwindows for growth and survival change during early life and are wider for eggs (5–17 �C) than in young(8- to 12-mm) early feeding larvae (5–12 �C). As larvae become able to successfully capture larger prey,thermal windows expand to include warmer waters. For example, 12- to 16-mm larvae can grow well at16 �C and larger, transitional-larvae and early juveniles display the highest rates of feeding and growth at�18–22 �C. Gaps in knowledge are identified including the need for additional laboratory studies on thephysiology and behavior of larvae (studies that will be particularly critical for biophysical modeling activ-ities) and research addressing the role of overwinter survival as a factor shaping phenology and settinglimits on the productivity of this species in areas located at the northern limits of its latitudinal range(such as the Baltic Sea). Based on stage- and temperature-specific mortality and growth potential of earlylife stages, our analysis suggests that young-of-the year sprat would benefit from inhabiting warmer,near-shore environments rather than the deeper-water spawning grounds such as the Bornholm Basin(central Baltic Sea). Utilization of warmer, nearshore waters (or a general increase in Baltic Sea temper-atures) is expected to accelerate growth rates but also enhance the possibility for density-dependent reg-ulation of recruitment (e.g., top-down control of zooplankton resources) acting during the late-larval andjuvenile stages, particularly when sprat stocks are at high levels.

� 2012 Elsevier Ltd. All rights reserved.

1. Sprat and the German GLOBEC program

The European sprat (Sprattus sprattus) was a main target speciesof the German GLOBEC program that investigated the trophody-namic structure and function of the Baltic and North Seas. Spratwas chosen as a research focus for three primary reasons. First,in some ecosystems sprat plays a prominent trophodynamic role

by exerting both top-down control on zooplankton and being anabundant prey resource for piscivores (wasp-waist species, e.g.Cury et al., 2000). Second, decadal trends in the abundance of theBaltic sprat stock were part of a profound regime shift from anAtlantic cod (Gadus morhua) to a clupeid-dominated system thatimpacted almost all trophic levels (Alheit et al., 2005; Möllmannet al., 2009) and appeared to be tightly coupled to physical (cli-mate) forcing. Third, the species is abundant within both the Balticand North Seas. Examining sprat in both systems provided theopportunity to compare and contrast its role in different food websand to determine if its life history strategy was flexible, allowing itto succeed in the vastly different conditions of physical forcing inthe Baltic and North Sea ecosystems.

Sprat is a small-bodied, pelagic schooling, zooplanktivorousclupeid that is distributed over a broad geographical range. InEuropean waters it occurs from the Black and Mediterranean Seas

0079-6611/$ - see front matter � 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.pocean.2012.04.013

⇑ Corresponding author.E-mail address: [email protected] (M.A. Peck).

1 Present address: School of Marine and Atmospheric Sciences, Stony BrookUniversity, Stony Brook, NY 11794-5000, USA.

2 Present address: Johann Heinrich von Thünen Institute, Federal Research Instituteof Rural Areas, Forestry and Fisheries (vTI) – Institute for Sea Fisheries, Palmaille 9,22767 Hamburg, Germany.

3 Present address: Sustainable Fishery, Department of Economics, University ofKiel, Wilhelm-Seelig Platz 1, 24118 Kiel, Germany.

Progress in Oceanography 103 (2012) 42–57

Contents lists available at SciVerse ScienceDirect

Progress in Oceanography

journal homepage: www.elsevier .com/locate /pocean

Author's personal copy

in the south to the European Atlantic shelf, including the North andBaltic Seas (Muus and Nielsen, 1999). It tolerates a wide range ofsalinities and is abundant in estuarine habitats. It is consideredan r-selected species (MacArthur and Wilson, 1967) because ofits short lifespan, early reproduction, and low biomass. It rarelyreaches an age >5 years (Bailey, 1980) or a length >16 cm (White-head, 1985). It is a multiple batch spawner, producing up to 10 eggbatches throughout the spawning season in some areas (Georgeand Alheit, 1987). Spawning occurs in both coastal and offshorewaters (Whitehead, 1985) and the time of peak spawning, relativefecundity, and batch fecundity vary significantly between yearsand regions (Alheit et al., 1987). Adults are generally mature at2 years though some individuals may spawn at 1 years (Bailey,1980). Genetic differences exist among sprat populations in Euro-pean waters (Debes et al., 2008; Limborg et al., 2009) and sub-spe-cies have been recognized.

Previous studies on sprat identified temperature as a key factoraffecting the population dynamics of this species at different lati-tudes. This is noted in both the scheduling of life-history eventsand inter-annual changes in the productivity of different spratstocks. For example, in the North Sea, peak spawning occurs inthe spring and early summer (Wahl and Alheit, 1988) while insouthern European waters (e.g., Adriatic Sea), sprat general spawnsduring the winter months (October–April) with peak spawning inNovember/December at water temperatures between 9 and 14 �C(Dulcic, 1998). MacKenzie and Köster (2004) examined recruit-ment strength at different water temperatures for the Baltic Seaand Black Sea stocks of sprat. Their inter-stock comparison sug-gested that recruitment was highest at intermediate water temper-atures experienced during spawning (5.0 and 9.0 �C) but tended tobe lower at colder (<3 �C) and warmer (P11 �C) waters (Fig. 1).Changes in a fitness trait (e.g., growth rate or fecundity) with anenvironmental factor often indicate changes in thermal optimaamong either (i) populations of a species inhabiting different lati-tudes, and/or (ii) life stages within a species. Both types of changescontribute to the scheduling of life-history events and define limitsto the characteristics of suitable habitats. These are discussed forsprat in subsequent sections.

This review summarizes available literature regarding the eco-physiology of sprat with an emphasis on progress made in the Ger-man GLOBEC program to understand how environmental factorsinfluence the life-history strategy and vital rates of this species inthe North and Baltic Seas. It compliments co-submissions on pro-

cess-oriented and modeling studies conducted on this species (Hin-richsen et al., submitted for publication; Voss et al., submitted forpublication) and on sprat prey-field dynamics (Schulz et al., sub-mitted for publication). This review discusses available ecophysio-logical data, distinguishing among studies conducted on the earliestlife stages (eggs, yolk sac and first-feeding larvae) and exogenouslyfeeding life-stages (larvae, transitional larvae, juveniles and adults).The impacts of abiotic factors (particularly temperature) on rates offeeding, growth, and reproduction are reviewed to understand thephysiological constraints that help shape the life-history strategyof this species in different regions. The focus is on physiologicaland life history attributes relevant for the ongoing developmentof biophysical, individual-based models of early life stages (e.g.,Peck and Hufnagl, 2012) and models that attempt to close the lifecycle (Rose et al., 2010). This review identifies gaps in knowledgeto recommend areas requiring future research.

2. Endogenous and mixed-feeding period

2.1. Eggs and yolk sac larvae

Sprat spawns pelagic eggs that are buoyant at different waterdepths in different systems due to salinity effects on ambient den-sity. In marine waters such as the North and Mediterranean Seas,eggs remain in surface layers but in the Baltic, eggs sink belowthe low salinity (�5–7 psu) surface waters through the thermo-cline to the halocline (�6–15 psu) located at intermediate waterdepths of 30–60 m (Wieland and Zuzarte, 1991). Research on sprateggs has focused on the impacts of water temperature and, to a les-ser extent, salinity on development rate and survival. Oxygen alsoplays an important role for egg survival in the Baltic Sea but effectshave not been examined in the laboratory. A number of studieshave examined the survival and developmental rate of sprat eggsincubated at temperatures between 1 and 20 �C (Thompson et al.,1981; Nissling, 2004; Kanstinger, 2007; Petereit et al., 2008).Thompson et al. (1981) incubated sprat eggs between 4.5 and20.0 �C and reported similar levels of survival between 4.5 and18.0 �C but indicated that, due to relatively early developmentalstage-at-hatch, larvae at >17 �C would likely not survive in the wild(Fig. 2a). Working with Baltic sprat eggs, Nissling (2004) reportedrelatively high (66–69%) viable hatch at temperatures between 5and 13 �C but low percentage hatch at <3 �C (0.2–26%) (Fig. 2a).In these studies, standard length (SL)-at-hatch was 3.3–3.5 mmand relatively similar at all temperatures <17 �C as was the SL atyolk sac absorption (4.9–5.6-mm SL) (Alshut, 1988b; Kanstinger,2007; Petereit et al., 2008). Several additional experiments havebeen conducted within the framework of GLOBEC Germany byPetereit (2008) on egg survival and developmental rate of eggsand yolk sac larvae that tend to confirm the results of previousstudies. Petereit et al. (2008) incubated eggs at 1.8–16 �C andfound the highest survival (percentage hatch) at 8 and 10 �C but,in contrast to other studies, no eggs survived when incubated attemperatures >14.7 �C. Similar incubation experiments performedon sprat eggs collected from the Adriatic Sea indicated high sur-vival (83–100%) at 11 different temperatures between 5 and19 �C (Petereit, 2008) which agrees with the results of previousstudies conducted on sprat eggs from the southern North Seaand English Channel (Thompson et al., 1981).

2.2. First-feeding larvae

One of the most important capabilities acquired during theearly life of larvae is the successful transition to exogenous feeding.Depending upon water temperature, sprat eye pigmentation oc-curs between 3 and 16 d post-hatch (dph) and jaw development

Fig. 1. Water temperature at the time of spawning in relation to sprat (Spratussprattus) recruitment in the Baltic Sea (squares, r2 adj = 0.28, p = 0.003) and BlackSea (circles, r2 adj = 0.35, p = 0.0002). A direct comparison of these two populationssuggests a temperature ‘‘optimum’’ for recruitment at spawning temperaturesbetween 5 and 9 �C. These data were digitized from MacKenzie and Köster (2004,their Fig. 2, p. 789). Black Sea data were compiled and analyzed by Daskalov (1999).

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 43

Author's personal copy

and mouth opening occur �48 and 72 h later (Nissling, 2004;Kanstinger, 2007). The ‘‘window of opportunity’’ for successfulfirst-feeding can thus be defined as the duration of time betweeneye pigmentation/mouth opening and the exhaustion of yolk re-serves or starvation mortality (Fig. 2b). A summary of measure-ments made in various studies on sprat yolk sac larvae suggeststhat the duration of time between these events can be 2–3 timeslonger at 8–10 �C (�11 d) compared to lower and higher tempera-tures (e.g., �6 and 4 d at 4 and 14 �C, respectively). This non-lineareffect of temperature on the window was due to starvation mortal-ity occurring more rapidly at 3 �C (�20 d) compared to 6 �C (23 d).At 14 �C mortality due to starvation occurred within 4 d (Petereitet al., 2008). Different populations of sprat may differ in the tem-perature-specific duration of this window. For example, the dura-tion of this window is slightly longer at the same, intermediatetemperatures in Baltic compared to Adriatic sprat (Petereit, 2008).

The results of these various studies on eggs and yolk sac larvaesuggest that temperatures between 7 and 13 �C support higher sur-vival of endogenously feeding stages. Temperatures between 5–7 �Cand 13–17 �C are sub-optimal and those in the ranges of 1–3 �C and

17–20 �C result in high mortality. Temperatures when spawning oc-curs have likely been set by these thermal boundaries. At constanttemperatures between 5 and 13 �C, the combined data in four stud-ies on eggs and yolk sac larvae (Thompson et al., 1981; Nissling,2004; Kanstinger, 2007; Petereit et al., 2008) indicated that theduration of the endogenous feeding period is 135 ± 3 degree-days(�C d) after which the larva is�5.5 mm SL and must initiate feeding.

3. Exogenously feeding life-stages

3.1. Ontogenetic changes in morphology

Different life-stages of fish can be identified by morphometicdifferences in relative body shape (Fuiman, 1983) and changes inthe slope, b, of the mass and length relationship (DM = a � SLb),can denote important changes in morphology associated withlife-history events (e.g., metamorphosis), changes in behavior(e.g., changes in swimming modes), and growth allocation strate-gies of energy acquired via feeding (Osse and van den Boogaart,1995) relationship between dry mass (DM) and SL. Unique charac-teristics of developmental stages often are expressed in changes inthe slope of the mass on length relationship. For example, in earlylife stages of fish, the slope (b) often differs (and is more variable)than that for large juveniles and adults (e.g., Pepin, 1995; Pecket al., 2005). The relationship between DM and SL for sprat hasbeen reported in a number of studies (Shields, 1989; Coull et al.,1989; Safran, 1992; Peck et al., 2005). Peck et al. (2005) describedchanges in sprat morphology by employing a segmented regres-sion having a slope of 5.0 for individuals of 5.5–44 mm SL butdeclining to 3.4 for larger sprat. These results agree well with(and explained the discrepancy between) slope values previouslyreported for 25–39 mm SL sprat (b = 5.6; Shields, 1989) and theslopes reported for larger YOY and adults (e.g. b = 2.83–3.47; Coullet al., 1989; Safran, 1992). Statistical models allowing for gradualrather than abrupt (segmented regression) changes in the slopemight be biologically more relevant (Peck et al., 2005). Indeed, are-analysis of those and additional morphometric data collectedfrom the Baltic Sea indicated that the scaling of dry mass andlength appears to change gradually (and non-linearly) withincreasing body size (Fig. 3a).

These changes in the slope (b-values) of the mass-length rela-tionship coincide with important life-history events in sprat. Forexample, the most rapid allocation of energy to dry mass per unitlength (highest b-value) occurs at body sizes between 14 to 18-mmSL (Fig. 3b). These sizes are most likely associated with the onset ofschooling in this species. We infer this from indirect evidence, spe-cifically that large numbers of 14-mm SL sprat were found withinthe guts of North Sea horse mackerel (Trachurus trachurus) (Matth-ias Bernreuther, personal observation). Within this range in stan-dard lengths, there is a rapid expansion of the prey niche breadth(see Section 3.2) and protein-specific growth rates (see Sec-tion 3.4.1). The rapid increase in DM per unit SL during this periodis likely due to rapid increases in the size (hypertrophy) and/ornumber (hyperplasia) of swimming muscle. It should be noted thatdifferences in the size and number of muscle fibers at differentbody sizes have not been examined in sprat and that swimmingmuscle development may also depend upon temperature basedupon the findings for larvae of coregonid and other clupeid species(Vieira and Johnston, 1992; Hanel et al., 1996). A second change inb-values can be noted during the metamorphosis of transitionallarvae into juveniles at 35–55-mm SL, after which point the adultbody form is obtained. The slope was also relatively high (4.3) atbody sizes of 60–80-mm SL, sizes obtained at the end of the firstgrowing season when sprat at relatively high latitudes (BalticSea) begin building somatic energy reserves in preparation for

Fig. 2. Effect of water temperature on eggs and yolk sac larvae of sprat (Spratussprattus). Combined data from two studies measuring the percentage survival ofsprat eggs incubated at different, constant temperatures (A). The function predictslittle change in relative survival of eggs between 5 and 16 �C. Time required for eyepigmentation and death (after yolk sac absorption) of unfed larvae in relation totemperature (data from three separate studies) (B). The stippled area represents anestimate of the ‘‘window of opportunity’’ for successful first-feeding (between eyepigmentation and death).

44 M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57

Author's personal copy

overwintering (see Section 3.4.3). Finally, b-values for adults exhi-bit seasonal changes in relation to changes in energy partitioningand allocation to growth in either length, mass, lipid or protein(see Section 3.4.3). The allocation strategy in the Baltic Sea in re-sponse to the timing of overwintering, spawning, and intensivefeeding periods is discussed more thoroughly below.

In summary, based on changes in growth allocation betweenmass and length and inferences from field observations (predatorgut contents), six life stages or life-history events can be identifiedthat occur after the egg and yolk-sac larval phases. For Baltic Seasprat, these include: (i) exogenously feeding but non-schooling lar-vae from 5 to 14-mm SL, (ii) likely onset of schooling behavior from14 to 18 mm-SL, (iii) a ‘‘transitional-larval’’ life stage from 18 to 35-mm SL, (iv) a period of late-larval/juvenile metamorphosis occurringat 35–55 mm-SL, (v) a juvenile growth phase from 55 to 90 mm-SL,and (vi) adult fish that exhibit seasonal energy allocation to somaticand gonadal (reproductive) growth starting at 100 mm-SL.

3.2. Prey field, diet, and foraging

Several field studies have examined the diets of larval sprat inthe Baltic (Wosnitza, 1974; Grauman and Yula, 1989; Arrhenius,

1996; Voss et al., 2003; Dickmann et al., 2007) and other ecosys-tems (Voss et al., 2009). The percentage feeding incidence, basedupon the numbers of larvae captured with food in their guts, wasreported to be low in the smallest larvae but increased rapidly withincreasing body size. Unfortunately, few laboratory studies havereared sprat beyond the yolk sac stage so observations under con-trolled conditions are rare. Kanstinger (2007) reared exogenouslyfeeding sprat larvae at 7, 10 and 13 �C in the presence of a nanofla-gellate (Rhodomonas baltica, 6–8 lm), a phagotrophic dinoflagel-late (Oxyrrhis marina, 12–25 lm) and copepod nauplii (Acartiatonsa, 125–175 lm). Dinoflagellates were found in the guts of alllarvae sampled at 8, 13, and 14 dph at 13, 10, and 7 �C, respectively,while copepod nauplii were observed in at least 25% of the guts 2–4 d later (10, 16, and 18 dph at 13, 10, and 7 �C, respectively). Theseobservations agree with the theoretical window of opportunity forfirst feeding (Fig. 3b). In situ collections often report unidentifiedmicroplankton in the guts of 5.5–10.5-mm sprat larvae (Vosset al., 2003; Dickmann et al., 2007). It is unknown whether thepresence of microplankton in guts resulted from active foragingor passive ‘‘drinking’’ and whether (or how much) larvae benefitfrom consuming this potential food source (e.g., whether this preyis digested and assimilated). However, diet studies have revealed

Fig. 3. Dry mass (DM) in relation to standard length (SL) for sprat (Spratus sprattus) larvae, juveniles, and adults, and changes in the allometric scaling of MD and SL with bodysize/life stage. Combined data set of MD and SL measurements made at different life stages were from Peck et al., 2004; Günther, 2008; J.-P. Herrmann, unpublished data(Panel A). Changes in the mean(±SE) slope (b) of the DM–SL relationship (DM = a � SLb) calculated for small ranges in lengths (shown as horizontal bars at each point) (Panel B).Fish length associated with schooling (i), late-larval to juvenile metamorphosis (ii) and juvenile to adult metamorphosis (iii) are als indicated.

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 45

Author's personal copy

that the amount of microplankton in sprat guts varied seasonally inBaltic Sea (Bornholm Basin) and tended to be positively related tofield estimates of survival (Voss et al., 2003; Dickmann et al., 2007).This suggests that microplankton may be necessary but not suffi-cient for high rates of survival of young larvae. These laboratoryand field results for young sprat larvae agree with recent labora-tory experiments indicating that incidental (or direct) feeding onprotists can expand the window of first feeding in sprat (Kanstin-ger, unpublished data) and Baltic cod (G. morhua) (Overton et al.,2010). These findings for sprat agree with observations by Pepinand Dower (2007) on the trophic position of larvae of other marinefish species and diet studies of the larvae of seven species in theNorth Sea (de Figuiredo et al., 2007) which suggest that compo-nents of the microbial loop (algae and heterotrophic protists) canbe important prey items.

The increase in the variance in prey sizes consumed (increase inniche breadth) with increasing larval size is a common feature insome marine fish larvae (Pepin and Penney, 1997) and is clearlyobserved in the gut contents of larval sprat. As sprat larvae increasein size, prey size also increases and, based on an analysis of com-bined data reported in different studies on Baltic sprat (Vosset al., 2003; Dickmann et al., 2007), prey size increases most rap-idly between 10 and 15 mm-SL (Fig. 4). At body sizes >15 mm-SL(including sizes of juveniles and adults not shown in Fig. 4), themaximum prey size changes relatively little (Last, 1987; Bernreu-ther, 2007). This initial increase in prey size (and increased vari-ance that leads to increased prey niche breadth) with increasingsprat size appears to be an important feature involved in growthpotential and patterns of growth with respect to temperature (dis-cussed below). The GLOBEC field sampling confirmed the impor-tance of copepods (Acartia sp.) in the diet of sprat larvae inrecent years (Voss et al., 2003; Dickmann et al., 2007). Seasonalcomparison of prey fields suggested that cladocerans might beimportant for survival of sprat larvae in summer, when cladoceranabundance reaches its peak.

Previous investigations, conducted during the early 1990s, indi-cated that Baltic sprat larvae performed diel vertical migrations(Voss et al., 2007). Interestingly, diel vertical migrations were notobserved during the German-GLOBEC study period (2002–2005).More recent (2007) field sampling in the southeastern Baltic(55�N, 19.25�E) confirm these decadal differences, particularly forsmall larvae (Karaseva and Ivanovich, 2010). The lack of diel migra-

tion in recent years is hypothesized to be due to changes in theabundance of copepod species which have different depth prefer-ences. Specifically, the abundances of Acartia and Temora specieswhich prefer surface waters has increased whereas the abundanceof Pseudocalanus which inhabits deeper depths has decreased (Vosset al., 2007; Schulz et al., submitted for publication). Biophysicalmodeling results of larval feeding and growth that included thesedecadal changes in prey fields and water temperatures suggestedfitness benefits related to the change in larval foraging behavior(Hinrichsen et al., 2010a). This is an example of how larval fishbehavior and life-history strategy adapt to changes in prevailingenvironmental conditions within an ecosystem. Regardless of thecause for the observed change in behavior, the lack of verticalmigration observed during the GLOBEC field studies is not only ex-pected to have consequences for survival but also profound influ-ences on larval transport (Hinrichsen et al., 2005, submitted forpublication).

The diet of juvenile and adult sprat life stages has been welldocumented for a variety of ecosystems (e.g., De Silva, 1973;Arrhenius and Hansson, 1993; Last, 1987; Cardinale et al., 2002;MacKenzie and Köster, 2004). A number of studies have examinedsprat diets in the Baltic Sea (Shvetsov et al., 1983; Szypula, 1985;Rudstam et al., 1992; Arrhenius, 1996; Szypula et al., 1997; Korni-lovs et al., 2001; Casini et al., 2004; Möllmann et al., 2004) due tothe important role of Baltic sprat and herring (Clupea harengus) inthe top-down regulation of zooplankton (Möllmann and Köster,2002; Casini et al., 2006). Studies have investigated daily feedingrhythms (Shvetsov et al., 1983) and compared diets among yearsand/or seasons (e.g., Szypula, 1985; Rudstam et al., 1992; Szypulaet al., 1997; Kornilovs et al., 2001; Möllmann et al., 2004). The mostrecent study indicated that the copepods Temora longicornis, Pseud-ocalanus acuspes, and Acartia spp. along with the cladocerans Evad-ne nordmanni, Bosmina longispina maritima, and Podon spp were themost common prey in the guts of large juvenile and adult sprat(Bernreuther, 2007). In one study examining diel feeding patterns,sprat strongly selected T. longicornis during the day and mainly Po-don spp. at night (Bernreuther, 2007). The reason for the diel shiftin prey selection was unknown. Adult copepods and older copepo-dites (C4–5) were usually selected and younger (smaller) develop-mental stages avoided. However, at certain times of the day,smaller stages were positively selected. For example, Temoracopepodites (c1–3) were positively selected between 19:30 and

Fig. 4. Sizes of prey items found in 1826 larval sprat (Spratus sprattus) guts in relation to larval standard length (mm). Data are for larvae collected in the Bornholm Basin,Baltic Sea and include those from Voss et al. (2003) and Dickmann et al. (2007).

46 M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57

Author's personal copy

01:30 (near and after sunset). During many German GLOBECcruises in the Baltic Sea, copepod nauplii were often abundant inzooplankton samples but rarely encountered in stomachs of largejuvenile and adult sprat (Bernreuther, 2007).

Large juveniles and adults of many clupeid species includingEuropean sardine (Sardina pilchardus), European anchovy (Engraulisencrasicolus), Atlantic herring, and Atlantic menhaden (Brevoortiatyrannus) employ either filter or particulate feeding dependingupon the size and concentration of prey items (Durbin and Durbin,1975; Garrido et al., 2007). However, sprat is an obligate particu-late feeder and its growth performance in nature might be stronglyinfluenced by variability in the abundance of larger (optimal) sizesof prey (discussed below). The relatively fixed rate of prey ‘‘snatch-ing’’ (rapid opening and closing of the mouth while swimmingthrough a patch of zooplankton) by juvenile and adult sprat hasimplications for the outcome of intra-specific and inter-specificcompetition for food. Intra-specific competitive behavior (in-creased swimming speeds) was observed when sprat schools wereprovided limited prey resources (M.A. Peck, unpublished data) andevidence from field surveys of age-0 juveniles (Baumann et al.,2007) and adults (J.-P. Herrmann, unpublished data) indicated thatboth life-stages may suffer from food limitation in the Baltic Sea. Itis speculated that the negative impacts of prey limitation would bemost severe for the largest adults that have similar maximum ratesof prey snatching (J.P. Herrmann, personal observation), and max-imum prey sizes, but have higher (absolute) daily energy require-ments compared to smaller adults.

The importance of feeding mode to sprat may be most evidentwhen inter-specific competition occurs with other small pelagicfishes and (potentially) invertebrate competitors such as the cteno-phore Mnemiopsis leidyi which has become established in the BalticSea (Haslob et al., 2007). European anchovy, sardine, herring nowco-occur with sprat in the North Sea. Gut content analyses of larvaeof sprat and sardine that co-occurred in the German Bight indi-cated a high degree of diet overlap but spatial differences in thedistribution of the two species minimized any potential competi-tion for prey (Voss et al., 2009). However, competition of juvenileand adult stages may be more intense due to their increase feedingabilities. In these life stages, the ability to switch between filter-and particulate feeding modes and to utilize a wider range in preysizes/types (e.g., see Raab et al., 2011) would appear to provide sar-dine, anchovy, and Atlantic herring a competitive advantage oversprat if these species experience poor feeding conditions, e.g. lowzooplankton concentrations and/or prey fields characterized byrelatively small zooplanktors). Furthermore, environmental ‘‘loop-holes’’ in predation (Bakun and Broad, 2003) may become availableto species that can utilize less productive environments that sup-port small populations (low concentrations) of relatively large zoo-plankton. Such environmental loopholes would likely beunavailable to sprat due to its inflexible feeding mode.

3.3. Rates of food consumption and temperature

3.3.1. Larval feedingWhen fish encounter high prey concentrations, daily rates of

feeding are, to a large extent, determined by the rate of gut evacu-ation, gut volume, and availability of sufficient light for visual for-aging. Gut evacuation rates can be qualitatively assessed for larvalsprat based upon the progressive decrease with time in gut con-tents of larvae sampled at the end of the day and at night. This ‘‘dielmethod’’ was applied to larval sprat sampled during three cruisesbetween late May and July 2003 in the Bornholm Basin, BalticSea (Dickmann et al., 2007). Sprat gut content data were pooledby 10-min time bins and feeding incidence (FI, number of fish withprey/total number of fish analyzed) and estimates of the mean drymass of food in each gut (lg, converted from counts and prey

length measurements) were examined in relation to the time ofcapture. Diel differences in feeding by sprat (n = 124, 13–16-mmSL) were evident (Fig. 5). The number of larvae with food in theguts was lower at dawn and dusk and no larvae had a gut with foodin the early morning and late in the day. The increase in both FI andgut content was variable during daylight hours. The former dis-played two (crepuscular) peaks whereas the latter generallyreached maximum values in the late afternoon in agreement withthe findings of Voss et al. (2003). At the end of the day, gut contentdeclined at a rate of �0.46(0.08) (mean (±SE)) h�1, an estimate sug-gesting that 13–16-mm SL sprat can empty their guts within about1.75–3.5 h (evacuation rates on the order of 40–50% h�1) duringnon-feeding periods with maximum mean dry mass of gut con-tents of �2.5% of larval dry mass (5 lg in a 200 lg DM larva,Fig. 5). These values agree with those measured in the larvae of avariety of different marine fish species (Peck and Daewel, 2007)and are particularly close to values obtained for herring larvae(e.g., Pedersen, 1984; Pepin and Penney, 2000). It should be notedthat gut evacuation rates depend upon water temperature. Caremust be taken when interpreting these field data since the larvaeof sprat are susceptible to damage when captured using standardsampling procedures (e.g., see Dänhardt and Temming, 2008)which might bias estimates if the degree of damage (or the occur-rence of stress-related gut evacuation) was related to tow durationor environmental conditions.

Estimates of feeding rates by sprat larvae are also available frommechanistic individual-based models (IBM’s) that include foragingand growth subroutines (Peck and Daewel, 2007; Daewel et al.,2008; Kühn et al., 2008). These IBM-based estimates of food con-sumption rate are derived from a balanced bioenergetics budget(including metabolic costs, assimilation efficiency of food, andother parameters) and otolith-based temperature-specific somaticgrowth rates of sprat larvae (discussed below). To balance ob-served growth rates, the maximum food consumption rate (CMAX,lg prey d�1) needed to change with body size and temperatureaccording to:

CMAX ¼ 1:315 � DM0:83 � 2:872T�15

10½ �

where T = water temperature (applied between 10 and 20 �C) andDM = larval dry mass (lg). At temperatures of 12 and 15 �C, CMAX

of a 200-lg DM sprat larva is predicted to be between 39% and53% larval DM d�1. The aforementioned in situ estimates of gut con-tents and (non-feeding) evacuation rates correspond to feedingrates of �20% larval DM d�1 during a 15-h photoperiod. In younglarvae, evacuation rates often increase 2-fold during feeding sug-gesting that modeled and in situ estimates of larval sprat food con-sumption rate agree (�40–45% DM d�1). Comparisons of rates offeeding by larvae of different marine fish species are difficult be-cause these rates are difficult to estimate and estimates vary widelyamong species (Houde and Schekter, 1983; Peck and Daewel, 2007).Houde and Zastrow (1993) reported a general, inter-specific rela-tionship describing the effect of temperature (T) on the rate of foodconsumption (C, d�1) by larvae (C = 0.0299T + 0.0389, see their Ta-ble 7) yielding food consumption rates between 40% and 49% DMd�1 for larvae at 12 and 15 �C.

3.3.2. Feeding rates in transitional-larvae and young juvenilesTransitional-larval and juvenile sprat were found to be rela-

tively amenable to laboratory work since they frequently occurin dense schools in shallow coastal waters and are little affectedby capture and transport to the laboratory. Direct measurementsof food consumption and growth have been made for 30 to 45-mm SL transitional-larval sprat in the laboratory (Peck et al.,2004). When tested at 14, 18, and 22 �C, the highest rates of foodconsumption were observed at 18 �C. At this temperature, food

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 47

Author's personal copy

consumption rates were between 36% and 44% fish DM d�1 (Fig. 6).At 22 and 14 �C, maximum feeding rates were �30% and 8% DMd�1, respectively. When groups of larval sprat at 18 �C were starvedfor 12 d and re-fed for an additional 12 d, no increased rates offeeding were observed in terms of the absolute number of zoo-plankton consumed indicating that 45% DM d�1 appears to be agood approximation of an upper limit (CMAX) for transitional larvaeand young juveniles of this species feeding during the summer inthe Baltic Sea. The CMAX by sprat (mean 50-mg DM) is essentially

the same as that (50% DM d�1) found for a variety of fishes at thatsame body size and water temperatures (e.g., Keckeis and Schi-emer, 1992, see their Fig. 7).

3.3.3. Feeding rates in juvenile and adult spratFeeding rates by large juvenile sprat (50–80-mm SL) have been

estimated in laboratory experiments examining gut fullness andgut evacuation rates. Using this method, rates of feeding have ex-ceeded 1.2% DM h�1 (Bernreuther et al., 2009) or �16% DM d�1

when using a 15-h feeding period. When feeding on live zooplank-ton (brine shrimp, Artemia sp., nauplii) in the laboratory, the meanwet weight of gut contents of large juvenile sprat could be up to 8%fish wet mass (WM) (J.P. Herrmann, personal observation). In con-trast, estimates of gut content mass have not exceeded 1.7% spratWM (Arrhenius, 1998). Typically, gut contents vary between 0.1%and 0.4% WM in Baltic Sea sprat (Bernreuther, 2007). In other sys-tems (e.g., Scottish west coast, Black Sea), the maximum gut con-tent weight in juvenile sprat appears to be between 0.4% and2.8% wet body mass for this species (De Silva, 1973; Sirotenkoand Sorokalit, 1979). Based upon a balanced bioenergetics budgetand known growth rates calculated from changes in the energycontent of age-0 juveniles captured monthly in the Bornholm Ba-sin, estimates of maximum feeding rates were 3.26% WM d�1

(�12% MD d�1) in June (Bernreuther, 2007). Hence, there is a largediscrepancy between observed gut contents, calculated feedingrates, and estimates of energy input needed for observed growthfor sprat in the Bornholm Basin. The discrepancy could be ex-plained by seasonal shifts in sprat feeding grounds from the deeperbasins to near-shore (shallower) waters. In that case, observedgrowth rates would be calculated for fish that returned to theBornholm Basin after feeding in another (more productive) regionthat was outside of the GLOBEC sampling area. Alternatively (or, inparallel), in situ feeding rates may be underestimated if fish evac-uate some of their gut content as a stress response during capture.

3.4. Growth rates and temperature

A number of approaches have been used within the GermanGLOBEC program to measure growth or growth potential of transi-tional larvae, juveniles, and adult sprat in the Baltic and North Seas.

Fig. 5. Mean dry mass of gut contents (GC, lg) and feeding incidence (FI, %) in relation to time of the day for 13- to 16-mm standard length sprat (Sprattus sprattus) larvaecaptured during three cruises between late May and July 2003 in the Bornholm Basin, Baltic Sea. Groups of larvae were pooled across cruises within 10-mm bins to generatethis figure (the number of larvae in each size bin is indicated at the top of the figure). FI = number of fish with prey (total number of fish analyzed)�1. GC was calculated fromprey lengths based upon prey species- and length-specific dry mass conversion factors. The times of sunrise (SR), sunset (SS) and civil twilight (CT) are indicated below theabscissa. The mean(±SE) slopes of the decrease in FI and GC (GC normalized to highest value) in relation to time were described by a slopes of �0.27(0.01) and �0.46(0.08),respectively (n = 4, p < 0.01). Details regarding larval sprat sampling were provided by Dickmann et al. (2007).

Fig. 6. Transitional larval and early juvenile sprat (Sprattus sprattus) growth rateversus food consumption rate in the laboratory at 14, 18 and 22 �C. Fish were fedbrine shrimp (Artemia sp.) nauplii for 10–12 d at each temperature. Regressions areshown for growth rate (GR, % fish dry mass d�1) versus food consumption (C, % fishdry mass d�1). Values for maintenance food consumption rates (CMAIN, C whereGR = 0) are also provided. See Peck et al. (2004) for details.

48 M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57

Author's personal copy

For example, otolith increment widths have been measured inyoung-of-the-year juveniles captured in autumn surveys in theBaltic (Baumann et al., 2006a,b, 2008) and North Seas (Baumannet al., 2009). Direct measurements of growth are available fromcontrolled laboratory studies performed on 25–45-mm SL transi-tional- (late-) larval sprat (Peck et al., 2004) and 55–70-mm SLjuveniles (J.-P. Herrmann, unpublished data). Growth rates of lar-ger juveniles and adults have been estimated from changes indry mass and length of fish captured in monthly cruises in the Bal-tic (J.-P. Herrmann, unpublished data). These studies combinedwith the research discussed here on early life-stages (eggs and lar-vae), provide an opportunity to assess the ontogeny of tempera-ture-dependent growth in sprat.

3.4.1. Growth rates of larvaeMost studies that investigate growth rates of marine fish focus

on mean growth rates of groups within cohorts, despite the factthat processes act upon the individual. An important distinctionshould be made between mean growth rates (e.g., those obtainedfor cohorts of larvae) and growth rates of individuals (e.g., larvaewithin those cohorts). Naturally, the former may be biased dueto the selective loss of either slow-growing individuals (e.g., dueto natural mortality) or fast-growing individuals (e.g., gear avoid-ance by relatively large larvae) and/or due to the mixing of cohortshaving different mean ages. For larval sprat, growth rates of indi-viduals are available from measurements of otolith incrementwidths and RNA-DNA ratios (Lee et al., 2006; Peck et al., 2007;Huwer, 2004; Hinrichsen et al., 2010b). These growth rates oftenreflect a high degree of growth variability among individual spratlarvae captured at the same time from the same station (e.g. coef-

ficient of variation of growth rates often >40%, Peck et al., 2007).With only a few exceptions (e.g., Voss et al., 2006; Hinrichsenet al., 2010b), inter-individual growth variability has been ignoredin most process-oriented field studies examining sprat growthrates. Therefore, for purposes of comparison we report meangrowth rates of groups in all subsequent sections. The vast major-ity of sprat growth data come from field studies because it has pro-ven difficult to rear larval sprat in the laboratory (e.g., Alshut,1988b; Shields, 1989; Kanstinger, 2007). The few estimates ofgrowth rate for laboratory-reared larval sprat tend to be lower thanthose from field studies (e.g., Fig. 7) even when compared togrowth rates of relatively small field-caught larvae in which issuesof gear avoidance can be ignored.

Mean growth rates of larval sprat have been determined fromotolith and biochemical (RNA-DNA ratios) methods applied to indi-viduals captured in the Baltic, North, Irish, and Adriatic Seas(Munk, 1993; Rè and Gonçalves, 1993; Dulcic, 1998; Valenzuelaand Vargas, 2002; Huwer, 2004; Holtappels, 2004; Lee et al.,2006; Voss et al., 2006; Daenhardt et al., 2007). Most estimatesof growth rates are for larvae <20 mm SL. There are indicationsfrom these studies that growth-temperature relationships varywith larval size. For example, biochemical-based growth rates for8–12-mm SL larvae in the Baltic Sea (Bornholm Basin, late Marchto early July 2002) declined at temperatures >9 �C (Fig. 8). In con-trast, no decline in protein-specific growth rates (or RNA-DNA ra-tios) was observed for larger (12- to 16-mm SL) larvae attemperatures up to 16 �C. Although we cannot separate the im-pacts of temperature from prey level in field studies of larvalgrowth, our biochemical measurements were made for sprat larvaefrom both size classes captured in multiple time periods which fur-

Fig. 7. Sprat (Sprattus sprattus) standard length (SL, mm) in relation to otolith ring count for larvae collected in six different field studies and SL in relation to larval age (days)reported in two laboratory studies. Three separate field studies collected larvae in the North Sea and one study collected larvae from the Baltic Sea. The back-calculatedlengths of sprat juveniles from the Baltic Sea are also shown for comparison. Regression equations for North and Baltic Sea larval data: circles, SL = 0.466 � ORC + 5.79,r2 = 0.849; triangles, SL = 0.405 � ORC + 7.79, r2 = 0.895; squares, SL = 0.444 � ORC + 6.38, r2 = 0.880; diamonds = 0.4414 � ORC + 6.311, r2 = 0.749). In all cases, regressions weresignificant (p < 0.05). Unlike field data, laboratory data include known ages of fish and include growth during and for 2 weeks after the yolk sac period at 14–15 �C.

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 49

Author's personal copy

ther supports our conclusion that these differences in growth po-tential between size classes at various temperatures reflect ontoge-netic changes in foraging efficiency. It may also correspond toontogenetic changes in metabolism which have not yet beenmeasured.

Due to increases in mouth gape and swimming ability, largersprat larvae can consume larger prey, and this increase in the spec-trum of available prey sizes would help support growth at highertemperatures. Smaller larvae would require a better ‘‘match’’ withprey (due to a smaller range in suitable prey sizes) to successfullyforage and grow. Moreover, smaller fish larvae are poorer swim-mers and have higher mass-specific metabolic costs at the sametemperature compared to larger larvae. Variation in growth dueto temperature and stage (size) observed in larval sprat may beindicative of more general, inter-specific trends since maximumprey size and mass-specific metabolic costs (e.g., routine respira-tion rates) generally increase and decrease, respectively, withincreasing body size (Houde, 1989). A general expectation wouldbe that larger and older larvae can exploit warmer habitats thansmaller and younger larvae. Indeed, young juveniles in many spe-cies often display the widest thermal tolerances (Pörtner and Peck,2010).

Otolith-based estimates of growth of North Sea sprat (ages 1–30 d post first-feeding) varied from 0.40 to 0.46 mm d�1 (Munk,1993; Rè and Gonçalves, 1993; Huwer, 2004) and are similar tothose (0.44 mm d�1) of Baltic sprat larvae (Hinrichsen et al.,2010a) (Fig. 7). Although many field studies have reported oto-lith-based mean growth rates for larval sprat cohorts, only Munk(1993) could detect significant differences in growth rates in rela-tion to hydrographic features (i.e. collection station position inrelation to a frontal zone in the North Sea). Only Munk (1993) em-ployed gear that caught larger sprat larvae and the association ofgrowth rates with frontal locations was only expressed in large lar-vae (>15 mm SL).

3.4.2. Growth rates of transitional larvae and juvenilesIn field-caught transitional-larval and juvenile sprat, the opti-

mal temperature for growth can be inferred from the widths ofotolith increments deposited during the first month after first feed-ing. These otolith studies have examined temperature and growthprior to and during the phase of transition between larval and late-

larval stages. The influence of temperature (T) on mean larval spratgrowth rate (GL, mm d�1) during the first 30 d post-feeding esti-mated from otolith-based growth rates from 0-group juveniles(Bornholm Basin, 2002 and 2003) and mean temperatures from ahydrodynamic model (Baumann et al., 2008) is described well by:

GL¼0:23 �T�0:005 �T2�1:69 ðr2¼0:90; n¼22; p<0:001Þ:

Growth rates were estimated to be 0.5 and 0.9 mm d�1 for fishin June and July at temperatures of 13–20.5 �C. The relationshipimplies Q10 values of 3.8 and 1.6 for changes in growth rates be-tween temperatures of 13–17 �C and 17–21 �C, respectively, and(by extrapolation) a temperature optimum at 22 �C. Also using oto-liths, Günther (2008) reported an exponential increase in growthrates of transitional-larval and young juvenile sprat captured atwater temperatures between 10 and 22 �C in nearshore areas ofthe western Baltic Sea. Again, estimates of temperature-dependentgrowth in wild fish are confounded because feeding history isunknown.

When transitional-larval and early juvenile sprat were providedad libitum rations of live zooplankton (Artemia nauplii) in short-term (4–5 d) laboratory feeding trials at 7, 11, 15, 18, and 22 �C,otolith increment widths and nucleic acids (RNA-DNA ratios)tended to reach their highest values between 18 and 22 �C(Fig. 9A and B) (Peck et al., 2004). These increases were most rapidbetween 10 and 15 �C and suggest that growth rates may not in-crease at temperatures >22 �C. This pattern of growth stabilization(or decline) at temperatures >22 �C agrees with the otolith growthhistories of late stage-larvae and early transitional larvae. The esti-mates from biochemical and otolith methods from short-termfeeding trials also agree well with the results of longer-term (12–14 d) feeding-growth trials. During those trials, maximum specificgrowth rates of transitional larvae were 2.5%, 11.5% and 7.5% DMd�1 at 15, 18 and 22 �C (see Fig. 6).

In somewhat larger juveniles, the physiological underpinning ofthe growth-temperature relationship has been examined by mea-suring rates of energy loss and energy gain at different tempera-tures. A strong dependence of feeding rates on temperature wasapparent from gastric evacuation experiments performed ongroups of juvenile (mean 63-mm TL, 283-mg DM) sprat fed Artemiasp. nauplii (Bernreuther et al., 2009). The rate of decrease in gutcontents changed with temperature in a non-linear manner. With

Fig. 8. Larval sprat (Sprattus sprattus) biochemical-based protein-specific growth rate in relation to water temperature at the time and location of field collection. Rates areshown for relatively small (8- to 12-mm standard length, SL) and large (12- to 16-mm SL) larvae. Boxes display the upper and lower 25th and 75th percentiles, whiskersdenote the 10th and 90th, circles represent measurements outside those ranges, and the thick line represents the mean value. Larvae were collected during five cruises in theBornholm Basin in late March to early July 2002. Multinet gear sampled larvae in discrete depths and temperature corresponds to the depth of capture. Symbols showingmeasurements for 8- to 12-mm SL larvae between 10 and 16 �C were shifted slightly so that they did not overlap with those for 12- to 16-mm SL larvae.

50 M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57

Author's personal copy

increasing temperature, the rate increased exponentially between10 and 15 �C, increased more slowly starting at approximately16 �C, and reached a maximum at 20 �C (Fig. 10A). The tempera-ture-dependent change in gut evacuation rates of 55-mm SL juve-niles agrees well with the temperature-dependent changes iningestion rates determined for smaller sprat transitional larvae(relatively low CMAX at 14 and 22 �C compared to 18 �C, describedpreviously). Rates of energy loss, measured via rates of O2 con-sumption by Meskendahl et al. (2010), appear to increase exponen-tially with increasing water temperature between 4 and 21 �C(Fig. 10B). Thus, the scope for growth (Brett, 1979) would appearto be relatively low at temperatures <10 �C and >22 �C, and maxi-mal between 18 �C and 20 �C in young juvenile sprat.

3.4.3. Spawning and growth of adultsSimilar to many other clupeiform species, sprat is an indetermi-

nate batch spawner that releases eggs over a prolonged period. In-tra- and interannual variability is expected in spawning seasonlength, batch fecundity, and batch frequency in all regions (Heid-rich, 1925; Alheit, 1988; Alshut, 1988a). Based upon the timingof spawning at different latitudes, spawning occurs between 6and 15 �C. In northern European waters (North and Baltic Seas),peak spawning occurs between May and August (Petrova, 1960;Wahl and Alheit, 1988) when water temperatures are commonlybetween 8 and 15 �C. In southern European waters (Adriatic Sea),sprat general spawns during the cooler time of the year (Octo-ber–April) with peak spawning in winter (November to December)at water temperatures between 9 and 14 �C (Dulcic, 1998). In allregions, however, the onset and duration of spawning may varydue to temperature and feeding conditions. Recent observationsin the Baltic detected spawning females as early as January (Ha-slob, 2011) and a second spawning peak observed in autumn

2003 was related to exceptional warm water masses during sum-mer in the Bornholm Basin (Kraus et al., 2004).

The batch fecundity (BF) of Baltic sprat was estimated via thehydrated oocyte method by Haslob et al. (2011). Those results indi-cated BF differed significantly among years and areas. For sprat inthe Bornholm Basin, BF increased linearly with increasing TLaccording to: BF = 413 � TL � 3510 (r2 = 0.43, n = 774, p < 0.05).Mean(±SE) size-specific BF varied from 86.0(±6.5) to 149(±4.5) eggs(g ovary-free WM)�1 for female sprat in the Bornholm Basin (Ha-slob et al., 2011) which is lower than sprat BF for the southernNorth Sea (413) and the Kiel Bight (232) (Alheit, 1988). Differencesin BF are likely driven by differences in environmental conditions(e.g., Petrova, 1960). In the case of Bornholm sprat, Haslob (2011)reported that absolute BF was positively related with water tem-perature (T, �C) during the pre-spawning period and fish size (TL,cm) according to:

BF ¼ 359:5ð�23:5Þ � TL � e�0:5�

ln T6:97ð�2:50Þ

� �1:46ð�0:55Þ

0@

1A

2

� 2753:2ð�243:5Þðr2 ¼ 0:70; n ¼ 179; p < 0:05Þ

where mean(±SE) parameter estimates are provided. During peakspawning, egg batches appeared to be released approximately every4 d (Haslob, 2011, spawning frequency estimated by macroscopicinspection of hydrated ovaries). These estimates agree well withprevious studies (Alekseev and Alekseeva, 2005; Kraus and Köster,2004).

In terms of seasonal growth patterns, the physiological studiesconducted on juveniles agree well with growth rates determinedfrom repeated sampling of juvenile and adult sprat in the Born-holm Basin of the Baltic Sea between February 2002 and May

Fig. 9. Growth proxies (otolith increment widths, Panel A; RNA-DNA ratio, Panel B)of 30- to 50-mm SL sprat in relation to water temperature from laboratory trials. In‘‘CMAX Trials’’, groups of sprat were maintained using ad libitum food rations for 5 dprior to measurements of growth proxies at each of the five temperatures. Meanvalues for 5–15 fish are shown. Two data points at 18 �C are mean values forsimilar-sized sprat maintained at ad libitum feeding conditions in 10-d ‘‘GrowthTrials’’ (see Fig. 6) and were not used to fit the regression lines. See Peck et al. (2004)for details.

Fig. 10. Rates of gut evacuation (A) and standard respiration (B) in relation totemperature for juvenile (55–70 mm standard length) sprat in the laboratory. Datain Panel A and B were from Bernreuther et al. (2009) and Meskendahl et al. (2010),respectively.

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 51

Author's personal copy

2004 (J.-P. Herrmann, unpublished data). Measurements of SL, DM,proximate composition (amounts of protein, lipid and energy), andcounts of otolith annuli revealed a more complete picture of theseasonal growth and energy allocation strategy of sprat in theBornholm Basin (Fig. 11A). For the different cohorts assessed (ageclasses), the seasonal allocation of energy to somatic and gonadalgrowth depends, to a large extent, upon the timing of spawning,the timing and availability of prey resources and physiological lim-itations imposed by seasonal changes in water temperature.

The seasonal changes in SL and DM (Fig. 11A) depict the life-his-tory strategy and growth dynamics of sprat in the Baltic, near thenorthern extreme of sprat’s geographical distribution. First, imma-ture stages of two cohorts (2002 and 2003) are first captured insummer or autumn which decrease in body size during their firstautumn and winter from �1.75 to �1.0 g DM. Inter-annual differ-ences in the growth of immature fish are also apparent. The2001 year class had small body sizes when first sampled (at age1) in April 2002. That 2001 year class likely matured at 2-year-olds.In spawning fish, the lowest DM was reached by approximatelyJune of each year (2002, 2003 and 2004) which marks the end ofthe main reproductive season. This is followed by a rapid butephemeral increase in DM by all year classes except the oldest(e.g., 6 year-old fish in 2003) which are lost from samples at thattime. The rapid increase in DM during the late spring and earlysummer is due mostly to increases in somatic protein content(Fig. 11B). This rapid increase in protein ended in mid to late sum-mer but DM continued to increase due to rapid deposition of lipidreserves later in summer (Fig. 11C). These lipid reserves are criticalsince, to a large extent, they fulfil the energy requirements of thisspecies during the overwinter period (when prey are scarce at thislatitude) and also fuel maturation and the onset of spawning activ-ity. The utilization of energy reserves is depicted by the decrease inDM observed throughout the overwintering and spawning periods.Feeding by sprat in the spring may help to extend the spawning

period by replenishing depleted energy reserves but this has notbeen directly examined.

Examples of this type of detailed seasonal growth data areavailable for a variety of marine fish species (e.g., Dygert, 1990;Smith and Paul, 1990; Jorgensen et al., 1997) but are uncommonfor clupeids. Unfortunately, no comparable data exist on the sea-sonal growth dynamics of sprat at lower latitudes where annualgrowth patterns would likely be different due to differences inthe phenology of prey resources and severity of the overwinteringperiod (see Conover, 1992). A recommendation stemming from theresults of field studies conducted in the German-GLOBEC programwould be to compare seasonal growth dynamics of sprat in the Bal-tic to those of conspecifics at lower latitudes. This would provide amore complete understanding of how abiotic and biotic factorsinteract to control ‘‘life-history scheduling’’ in this species.

4. Temperature-dependent life-history strategy in the Baltic Sea

This summary of knowledge on life stage- and temperature-specific growth patterns enables a more thorough understandingof the constraints placed on ‘‘life-cycle scheduling’’ of sprat, espe-cially in the Baltic Sea. Based on laboratory and field studies, thereappears to be ontogenetic changes in the range of temperaturesthat support growth with transitional-larval and early juvenilestages able to exploit warmer temperatures than eggs, larvae oradults (Fig. 12a). Theory suggests that young juveniles may be ableto exploit (grow well in) a larger range in temperatures comparedto either earlier (eggs/larvae) or later (mature adult) life stages(Pörtner and Farrell, 2008; Rijnsdorp et al., 2009; Pörtner and Peck,2010) and our summary for sprat supports this theory. In othertemperate marine fish species such as Atlantic cod (G. morhua)and sole (Solea solea), ontogenetic expansion and contraction oftolerable thermal environments have been reported for juveniles

Fig. 11. Phenology of growth of sprat (Sprattus sprattus) in the Bornholm Basin (Baltic Sea). Measurements include dry mass (A) total protein content (Panel B) and total lipidcontent (C) for each of seven sprat year classes (1997–2003). Basin-wide mean values are shown for each cohort on each date. Three periods within each year are recognized(different shades of gray): spawning phase, summer somatic growth phase, and an over-wintering phase. During the summer, growth is initially allocated to increase proteincontent (and length, not shown) and, subsequently, to a rapid increase in lipid to prepare individuals for over-wintering and gonadal maturation.

52 M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57

Author's personal copy

and spawning adult phases, respectively, and stem from increasesand decreases in aerobic performance capacity (Pörtner and Peck,2010).

There are three, main temperature-dependent constraints thatimpact the scheduling of life-history events for Baltic sprat: sur-vival of eggs, success of YOY overwintering and energy partitioningand gonadal maturation of adults (Fig. 12b). Successful spawning(spawning that produces viable offspring) would not occur untilwaters are P5 �C due to low survival of eggs incubated at coldertemperatures. Interestingly, temperatures <5 �C are activelyavoided by large juvenile and adult sprat (Stepputtis, 2006). Tem-peratures supporting successful development of early life stages ofsprat occur (at depth) for much of the year (April–November) butthe end of the spawning season is constrained by the time requiredby offspring to reach (juvenile) body sizes that can successfullyoverwinter. Young-of-the-year overwintering mortality, a commongrowth constraint in populations of freshwater and marine fishinhabiting temperate latitudes (Schultz et al., 1998; Gotceitaset al., 1999; Höök et al., 2007), has not been examined in sprat

but size requirements for successful overwintering can be inferredfrom in situ observations. Specifically, sprat consistently obtained amean size of P75 mm SL at the end of the first growing season dur-ing time periods when Baltic sprat year class recruitment indiceswhere quite high (ICES, 2009). The mean SL of YOY sprat capturedin November in 2002, 2003, 2005, and 2006 was 79.3, 78.5, 81.7and 73.7 mm, respectively (J.-P. Herrmann, unpublished data). Oto-lith-based growth histories of survivors (age-0 juveniles) caught inNovember 2002 (data from H. Baumann) indicated that about�100 d (mean ± range of 98 ± 7 d) were needed for growth from12.0 mm to 75 mm SL in that year. Given average water tempera-tures in the Bornholm Basin and based upon estimates of thedevelopment and growth rates of earlier life stages (see earlier sec-tions on eggs and yolk sac larvae), sprat would require �120–140 dafter fertilization to obtain this pre-winter size. Pre-recruits wouldnot be able to reach this pre-winter size if spawning took placeafter mid-July (see Fig. 12). The lack of eggs in surveys conductedin mid-July supports this sprat life-history constraint in the Born-holm Basin of the Baltic Sea (R. Voss, personal observation).

Fig. 12. Temperature-specific growth potential and life history scheduling of sprat (Sprattus sprattus) in the Baltic Sea. Relative impact of different water temperatures on thevital rates (rates of survival, feeding, and growth) of various life stages (Panel A). This conceptual diagram is based upon multiple field and laboratory studies discussedelsewhere in the text. The phenology and scheduling of various life history events for sprat in the Bornholm Basin (Panel B) was based upon mean, depth-specific watertemperatures for 2002 and 2003 (shown at the bottom). The first appearance of eggs, larvae, transitional larvae and juveniles is constrained by the availability of tolerable(warm) surface water temperatures (and the development rate of earlier life stages). The time period of final appearance of these life stages is constrained by temperature-dependent growth potential and size thresholds observed prior to overwintering that are potentially required for overwinter survival. Note, surface water temperatures aretoo warm for eggs and young larvae during the summer but cooler, optimal temperatures occur at depth. The phenology of spawning and protein and lipid growth (gray bars)are based upon from field measurements (see Fig. 11).

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 53

Author's personal copy

A combined analysis of otolith-based growth trajectories andhydrodynamic modeling (providing likely temperatures experi-enced) predicted that young-of-the year sprat always benefitedfrom relatively warm temperatures in the Bornholm Basin (Bau-mann et al., 2008). However, optimal temperatures predicted forgrowth in transitional-larval and juvenile sprat (�20–22 �C) occurfor only a short time period (e.g., �2 weeks in late July, 2003, seeFig. 12b) within the Bornholm Basin, one of the main spawninggrounds. From a physiological perspective, shallow inshore watersthat warm more than offshore areas would be more optimal areasfor rapid growth of transitional larvae and young juvenile sprat,provided that these areas have ample prey fields to supportgrowth, which may not always be the case (see Baumann et al.,2007). Finally, a certain amount of time is required for post-spawn-ing adults to forage and partition energy to somatic growth. Workis in progress to develop energy partitioning rules (e.g., what foodenergy is allocated to gonadal versus somatic growth) to modellife-history growth strategies (K. Andersen, DTU Aqua, personalcommunication).

Differences exist between life-history attributes of North Seaand Baltic Sea sprat that do not appear to be driven by physiolog-ical differences between these populations. In fact, most facets ofthe physiology appear similar between the Baltic and North Seapopulations (development rates of endogenously feeding lifestages, growth rates of young larvae, etc.). Although a recent anal-ysis utilizing nine microsatellite markers uncovered genetic differ-ences between these populations (Limborg et al., 2009), wepostulate that differences in life-history scheduling only reflect dif-ferences in seasonality in water temperature and prey production.In the North Sea, the sprat spawning season extends later in thesummer (sometimes through August) and, at the end of their firstgrowing season, sprat are 10–12 mm (�20%) smaller than theirBaltic conspecifics. We speculate that the severity of overwinteringconditions is an important factor regulating life-history schedulingin northern populations of sprat. A larger pre-winter size may berequired in the Baltic Sea compared to the North Sea due to theharsher environment experienced during the winter (and perhapsdue to the lower salinities) in the former system. However, over-wintering zooplankton populations, potential feeding and size-specific survival have not been assessed in either the North Seaor Baltic Sea. This lack of knowledge on overwintering dynamicsappears to exist for most small pelagic fishes that now exist inthe North and Baltic Seas (e.g., sprat, herring, European anchovy,European sardine). Ecophysiological constaints such as factorsaffecting overwintering survival are clearly a relevant topics to ad-dress in future research programs that attempt to go beyondmerely correlating changes in abundance and distribution to cli-mate indices (e.g., Alheit et al., 2012) but also attempt to revealthe mechanisms causing these fluctuations (e.g., Hufnagl and Peck,2011; Petitgas et al., 2012) in order to gain a cause-and-effectunderstanding that will aid in projections of future conditions.

5. Conclusions: From physiology to process knowledge

This study summarized existing knowledge on the ecology andgrowth physiology of sprat and highlighted recent knowledgegained during the German GLOBEC program. Considerable infor-mation now exists on the ecophysiology of most life stages of thisspecies. The review was restricted to aspects of feeding and growthphysiology and did not encompass a variety of topics (e.g., migra-tion behavior, stock structure, population demographics, mortalityrates, and recruitment patterns). Field and laboratory research al-lowed, among other things, the development of physiologically-based growth models including mechanistic individual-basedmodels (IBMs) of larval foraging and growth (e.g., Peck and Daewel,

2007). These IBM’s, when coupled to 3-D hydrodynamic models,are tools that are normally developed and applied towards theend of large-scale fisheries oceanographic research programs (likeGLOBEC-Germany) as a way of synthesizing and testing processknowledge. Larval sprat IBMs utilized in German GLOBEC haveexamined the impacts of prey field variability on survival andgrowth in the North Sea (Daewel et al., 2008; Kühn et al., 2008)and Baltic Sea (Hinrichsen et al., 2010a, submitted for publication).

Knowledge on stage-specific growth physiology is a prerequi-site for building robust process models that attempt to explain(in a mechanistic fashion) environmental constraints on the vitalrates and recruitment of marine fish populations and projectingclimate impacts (Pörtner and Farrell, 2008; Rijnsdorp et al., 2009;Pörtner and Peck, 2010). Based on our review, a number of gapsin knowledge were identified including processes related to (1)first-feeding prey and feeding success (potential role of the micro-bial loop), (2) ontogenetic development during the larval period(physiological changes permitting larvae to inhabit and grow wellin higher temperatures), and (3) late-stage juveniles (high rate offeeding and the role of overwinter mortality as a constraint tolife-history scheduling in northern areas). Such topics would begermane for future investigations on sprat and many other temper-ate marine fish species. Based on sequential correlations of theabundance of different life stages with recruitment, the transi-tional-larval/early juvenile phase was recognized as a critical per-iod in Baltic sprat (Köster et al., 2003; Baumann et al., 2006a; Vosset al., submitted for publication) but the mechanisms were notidentified. In terms of growth physiology of sprat in the BalticSea, it appears that cohorts produced relatively late in the spawn-ing season can contribute the bulk of survivors in some years (Vosset al., submitted for publication) and access of YOY juveniles to rel-atively warm, productive waters would be required to enhancegrowth and survival. On the other hand, at high stock sizes (whenjuveniles are very abundant), this life-history strategy might in-crease the potential for top-down control of zooplankton. This den-sity-dependent mechanism would reduce survival and recruitmentboth indirectly (by reducing growth rates making individuals morevulnerable to predation) and directly (via starvation prior to orduring the overwinter period).

Acknowledgements

The authors would like to thank all of the participants of theGerman GLOBEC program including the many laboratory assistantsand members of research vessel crews that helped collect the datathat were presented within this manuscript. GLOBEC Germany wasfunded by the German Federal Ministry for Education and Research(FKZ 03F0320E). Partial funding for this research was also receivedfrom the ‘‘FACTS’’ (Forage Fish Interactions, EU FP7, 244966) re-search program.

References

Alekseev, F.E., Alekseeva, E.I., 2005. Batch fecundity and daily egg production of theBaltic sprat Sprattus sprattus balticus (Clupeidae) from the southeastern part ofthe Baltic Sea. Journal of Ichthyology 45, 93–102.

Alheit, J., 1988. Reproductive biology of sprat (Sprattus sprattus): factorsdetermining annual egg production. Conseil International pour le Explorationde la Mer 44, 162–168.

Alheit, J., Wahl, E., Cibangir, B., 1987. Distribution, abundance, development rates,production and mortality of sprat eggs. International Council for theExploration of the Seas. ICES CM/H:45.

Alheit, J., Möllmann, C., Dutz, J., Kornilovs, G., Loewe, P., Mohrholz, V., Wasmund, N.,2005. Synchronous ecological regime shifts in the central Baltic and the NorthSea in the late 1980s. ICES Journal of Marine Science 62 (7), 1205–1215.

Alheit, J., Pohlmann, T., Casini, M., Greve, W., Hinrichs, R., Mathis, M., O’Driscoll, K.,Vorberg, R., Wagner, C., 2012. Climate variability drives anchovies and sardinesinto the North and Baltic Seas. Progress in Oceanography 96, 128–139.

54 M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57

Author's personal copy

Alshut, S., 1988a. Seasonal variations in length frequency and birthdate distributionof juvenile sprat (Sprattus sprattus). International Council for the Exploration ofthe Sea. ICES CM 1988/H:44.

Alshut, S., 1988b. Daily growth increments on otoliths of laboratory reared sprat,Sprattus sprattus L., larvae. Meeresforschung 32, 23–29.

Arrhenius, F., 1996. Diet composition and food selectivity of 0-group herring (Clupeaharengus L.) and sprat (Sprattus sprattus L.) in the northern Baltic Sea. ICESJournal of Marine Science 53, 701–712.

Arrhenius, F., 1998. Food intake and seasonal changes in energy content of youngBaltic Sea sprat (Sprattus sprattus L.). ICES Journal of Marine Science 55, 319–324.

Arrhenius, F., Hansson, S., 1993. Food consumption of larval, young and adultherring and sprat in the Baltic Sea. Marine Ecology Progress Series 96, 125–137.

Bailey, R.S., 1980. Problems in the management of short-lived pelagic fish asexemplified by North Sea sprat. Rapports et Proces-Verbaux des Reunions,Conseil International pour L’Exploration de la Mer 177, 477–488.

Bakun, A., Broad, K., 2003. Environmental ‘loopholes’ and fish population dynamics:comparative pattern recognition with focus on El Niño effects in the Pacific.Fisheries Oceanography 12, 458–473.

Baumann, H., Hinrichsen, H.-H., Voss, R., Stepputtis, D., Grygiel, W., Clausen, L.W.,Temming, A., 2006a. Linking growth- to environmental histories in centralBaltic young-of-the-year sprat, Sprattus sprattus: an approach based on otolithmicrostructure analysis and hydrodynamic modeling. Fisheries Oceanography15, 465–476.

Baumann, H., Gröhsler, T., Kornilovs, G., Makarchouk, A., Feldman, V., Temming, A.,2006b. Temperature-induced regional and temporal growth differences inBaltic young-of-the-year sprat, Sprattus sprattus. Marine Ecology Progress Series317, 225–236.

Baumann, H., Peck, M.A., Götze, E., Temming, A., 2007. Starving early juvenile sprat,Sprattus sprattus L., in Western Baltic coastal waters: evidence from combinedfield and laboratory observations in August/September 2003. Journal of FishBiology 70, 853–866.

Baumann, H., Voss, R., Hinrichsen, H.-H., Mohrholz, V., Schmidt, J.O., Temming, A.,2008. Investigating the selective survival of summer- over spring-sprat, Sprattussprattus, in the Baltic Sea. Fisheries Research 91, 1–14.

Baumann, H., Voss, R., Malzahn, A., Temming, A., 2009. The German Bight (NorthSea) is a nursery area for both locally and externally produced sprat juveniles.Journal of Sea Research 61, 234–243.

Bernreuther, M., 2007. Investigations on the feeding ecology of Baltic Sea herring(Clupea harengus L.) and sprat (Sprattus sprattus L.). PhD Thesis. Institute forHydrobiology and Fisheries Science, University of Hamburg, Hamburg,Germany, 184pp.

Bernreuther, M., Temming, A., Herrmann, J.-P., 2009. Gastric evacuation in sprat(Sprattus sprattus L.). Journal of Fish Biology 75, 1529–1541.

Brett, J.R., 1979. Environmental factors and growth. In: Hoar, W.S., Randall, D.J.,Brett, J.R. (Eds.), Fish Physiology, vol. VIII. Academic Press, London, pp. 599–667.

Cardinale, M., Casini, M., Arrhenius, F., 2002. The influence of biotic and abioticfactors on the growth of sprat (Sprattus sprattus) in the Baltic Sea. Aquatic LivingResources 15, 273–281.

Casini, M., Cardinale, M., Arrhenius, F., 2004. Feeding preferences of herring (Clupeaharengus) and sprat (Sprattus sprattus) in the southern Baltic Sea. ICES Journal ofMarine Science 61, 1267–1277.

Casini, M., Cardinale, M., Hjelm, J., 2006. Inter-annual variation in herring, Clupeaharengus, and sprat, Sprattus sprattus, condition in the central Baltic Sea: whatgives the tune? Oikos 112, 638–650.

Conover, D.O., 1992. Seasonality and the scheduling of life history at differentlatitudes. Journal of Fish Biology 41 (Suppl. sB), 161–178.

Coull, K.A., Jermyn, A.S., Newton, A.W., Henderson, G.I., Hall, W.B., 1989. Length/weight relationships for 88 species of fish encountered in the North Atlantic.Scottish Fisheries Research Reports 43, 1–80.

Cury, P., Bakun, A., Crawford, R.J., Jarre, A., Quinones, R.A., Shannon, L.J., Verheye,H.M., 2000. Small pelagics in upwelling systems: patterns of interaction andstructural changes in ‘‘wasp-waist’’ ecosystems. ICES Journal of Marine Science57, 603–618.

Daenhardt, A., Peck, M.A., Clemmesen, C., Temming, A., 2007. Depth-dependentnutritional condition of sprat larvae in the central Bornholm Basin. Baltic SeaMarine Ecology Progress Series 341, 217–228.

Daewel, U., Peck, M.A., Kühn, W., St. John, M.A., Alekseeva, I., Schrum, C., 2008.Coupling ecosystem and individual-based models to simulate the influence ofenvironmental variability on potential growth and survival of larval sprat(Sprattus sprattus L.) in the North Sea. Fisheries Oceanography 17, 333–351.

Dänhardt, A., Temming, A., 2008. Effects of haul duration on the physical conditionof sprat larvae. Aquatic Biology 1, 135–139.

Daskalov, G., 1999. Relating fish recruitment to stock biomass and physicalenvironment in the Black Sea using generalized additive models. FisheriesResearch 41, 1–23.

de Figuiredo, G.M., Nash, R.D.M., Montagnes, D.J.S., 2007. Do protozoa contributesignificantly to the diet of larval fish in the Irish Sea? Journal of the MarineBiological Association of the UK 87, 843–850.

De Silva, S.S., 1973. Food and feeding habits of the herring Clupea harengus and thesprat C. sprattus in inshore waters of the west coast of Scotland. Marine Biology20, 282–290.

Debes, P.V., Zachos, F.E., Hanel, R., 2008. Mitochondrial phylogeography of theEuropean sprat (Sprattus sprattus L., Clupeidae) reveals isolated climaticallyvulnerable populations in the Mediterranean Sea and range expansion in thenortheast Atlantic. Molecular Ecology 17, 3873–3888.

Dickmann, M., Möllmann, C., Voss, R., 2007. Feeding ecology of Central Baltic spratSprattus sprattus larvae in relation to zooplankton dynamics: implications forsurvival. Marine Ecology Progress Series 342, 277–289.

Dulcic, J., 1998. Larval growth of sprat, Sprattus sprattus phalericus, larvae in thenorthern Adriatic. Fisheries Research 36, 117–126.

Durbin, E., Durbin, A., 1975. Grazing rates of the Atlantic menhaden Brevoortiatyrannus as a function of particle size and concentration. Marine Biology 33,265–277.

Dygert, P., 1990. Seasonal changes in energy content and proximate compositionassociated with somaticgrowth and reproduction in a representative age–classof female English sole. Transactions of the American Fisheries Society 119, 791–801.

Fuiman, L.A., 1983. Growth gradients in fish larvae. Journal of Fish Biol. 23, 117–123.Garrido, S., Marçalo, A., Zwolinski, J., van der Lingen, C.D., 2007. Laboratory

investigations on the effect of prey size and concentration on the feedingbehaviour of Sardina pilchardus. Marine Ecology Progress Series 330, 189–199.

George, M.R., Alheit, J., 1987. Ovarian maturation cycle of sprat, Sprattus sprattus.International Council for the Exploration of the Sea. ICES CM/H 47.

Gotceitas, V., Methven, D.A., Sandy Fraser, S., Brown, J.A., 1999. Effects of body sizeand food ration on over-winter survival and growth of age-0 Atlantic cod, Gadusmorhua. Environmental Biology of Fishes 54, 413–420.

Grauman, G.B., Yula, E., 1989. The importance of abiotic and biotic factors in earlyontogenesis of cod and sprat. Rapports et procès-verbaux des réunions, Conseilpermanent international pour l’exploration de la mer 190, 207–210.

Günther, C., 2008. Untersuchungen zum Sommerwachstum juveniler Sprotten(Sprattus sprattus) in der westlichen Ostsee. University of Hamburg,Diplomarbeit, 89 pp.

Hanel, R., Karjalainen, J., Wieser, W., 1996. Growth of swimming muscles and itsmetabolic cost in larvae of whitefish at different temperatures. Journal of FishBiology 48, 937–951.

Haslob, H., 2011. Reproductive ecology of Baltic sprat and its application in stockassessment. Dissertation. University of Kiel, Kiel, Germany, 133 pp.

Haslob, H., Clemmesen, C., Schaber, M., Hinrichsen, H.-H., Schmidt, J.O., Voss, R.,Kraus, G., Köster, F.W., 2007. Invading Mnemiopsis leidyi as a potential threat toBaltic fish. Marine Ecology Progress Series 349, 303–306.

Haslob, H., Tomkiewicz, J., Hinrichsen, H.H., Kraus, G., 2011. Spatial and interannualvariability in Baltic sprat batch fecundity. Fisheries Research 110 (2), 289–297.

Heidrich, H., 1925. Über die Fortpflanzung von Clupea sprattus in der Kieler Bucht.Wissenschaftliche Meeresuntersuchungen 20, 1–47.

Hinrichsen, H.-H., Kraus, G., Voss, R., Stepputtis, D., Baumann, H., 2005. The generaldistribution pattern and mixing probability of Baltic sprat juvenile populations.Journal of Marine Systems 58, 52–66.

Hinrichsen, H.-H., Peck, M.A., Schmidt, J., Huwever, B., Voss, R., 2010a. Decadalchanges in the diel vertical migration behaviour of Baltic sprat larvae: causesand consequences. Limnology and Oceanography 55 (4), 1484–1498.

Hinrichsen, H.H., Voss, R., Huwer, B., Clemmesen, C., 2010b. Variability of larvalBaltic sprat (Sprattus sprattus L.) otolith growth: a modeling approachcombining spatially and temporally resolved biotic and abiotic environmentalkey variables. Fisheries Oceanography 19, 463–479.

Hinrichsen, H.-H., Kühn, W., Peck, M.A., Voss, R., submitted for publication. Theimpact of physical and biological factors on the drift and spatial distribution oflarval sprat: a comparison of the Baltic and North Seas. Progress inOceanography.

Holtappels, M., 2004. The nutritional condition of sprat and sardine larvae in thefrontal systems of the German Bight. M.Sc. Thesis. University of Rostock,Institute for Baltic Sea Research, Warnemünde, Germany, 87 pp.

Höök, T.O., Rutherford, E.S., Mason, D.M., Carter, G.S., 2007. Hatch dates, growth,survival, and overwinter mortality of age-0 alewives in Lake Michigan:implications for habitat-specific recruitment success. Transactions of theAmerican Fisheries Society 136, 1298–1312.

Houde, E.D., 1989. Comparative growth, mortality, and energetics of marine fishlarvae: temperature and implied latitudinal effects. US Fishery Bulletin 87, 471–495.

Houde, E.D., Schekter, R.C., 1983. Oxygen uptake and comparative energetics amongeggs and larvae of three subtropical marine fishes. Marine Biology 72, 283–293.

Houde, E.D., Zastrow, C.E., 1993. Ecosystem- and taxon-specific dynamic andenergetic properties of larval fish assemblages. Bulletin of Marine Science 53(2), 290–335.

Hufnagl, M., Peck, M.A., 2011. Physiological-based modelling of larval Atlanticherring (Clupea harengus) foraging and growth: insights on climate-driven lifehistory scheduling. ICES Journal of Marine Science 68 (6), 1170–1188.

Huwer, B., 2004. Larval growth of Sardina pilchardus and Sprattus sprattus in relationto frontal systems in the German Bight. Diploma Thesis. Institut fürMeereskunde, Christian-Albrechts-Universität zu Kiel, 108 p.

ICES, 2009. Report of the Baltic Fisheries Assessment Working Group (WGBFAS),22–28 April 2009. ICES Headquarters, Copenhagen. ICES CM 2009/ACOM:07,626 pp.

Jorgensen, E.H., Johansen, S.J.S., Jobling, M., 1997. Seasonal patterns of growth, lipiddeposition, and lipid depletion in anadromous Arctic charr. Journal of FishBiology 51, 312–326.

Kanstinger, P., 2007. Growth and condition of larval sprat (Sprattus sprattus): acombined laboratory and field investigation. University of Hamburg,Diplomarbeit, 86 pp.

Karaseva, E.M., Ivanovich, V.M., 2010. Vertical distribution of eggs and larvae of theBaltic Sprat Sprattus sprattus balticus (Clupeidae) in relation to seasonal anddiurnal variation. Journal of Ichthyology 50, 259–269.

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 55

Author's personal copy

Keckeis, H., Schiemer, F., 1992. Food consumption and growth of larvae andjuveniles of three cyprinid species at different food levels. EnvironmentalBiology of Fishes 33, 33–45.

Kornilovs, G., Sidrevics, L., Dippner, J.W., 2001. Fish and zooplankton interaction inthe Central Baltic Sea. ICES Journal of Marine Science 58, 579–588.

Köster, F.W., Hinrichsen, H.-H., Schnack, D., St. John, M.A., MacKenzie, B.R.,Tomkiewicz, J., Möllmann, C., Kraus, G., Plikshs, M., Makarchouk, A., Eero, A.,2003. Recruitment of Baltic cod and sprat stocks: identification of critical lifestages and incorporation of environmental variability into stock-recruitmentrelationships. Scientia Marina 67 (Suppl. 1), 129–154.

Kraus, G., Köster, F.W., 2004. Estimating Baltic sprat (Sprattus sprattus balticus S.)population size from egg production. Fisheries Research 69, 313–329.

Kraus, G., Mohrholz, V., Voss, R., Dickmann, M., Hinrichsen, H.H., Hermann, J.P.,2004. Consequences of summer inflow events on the reproduction cycle ofBaltic sprat. International Council for the Exploration of the Sea. ICES CM/L:19.

Kühn, W., Peck, M.A., Hinrichsen, H.-H., Daewel, U., Moll, A., Pohlmann, T., Stegert,C., Tamm, S., 2008. Defining habitats suitable for larval fish in the German Bight(southern North Sea): an IBM approach using spatially- and temporally-resolved, size-structured prey fields. Journal of Marine Systems 74, 329–342.

Last, J.M., 1987. The food of immature sprat (Sprattus sprattus (L.)) and herring(Clupea harengus L.) in coastal waters of the North Sea. Journal du ConseilInternational pour lExploration de la mer 44, 73–79.

Lee, O., Danilowicz, B.S., Dickey-Collas, M., 2006. Temporal and spatial variability ingrowth and condition of dab (Limanda limanda) and sprat (Sprattus sprattus)larvae in the Irish Sea. Fisheries Oceanography 15, 490–507.

Limborg, M.T., Pedersen, J.S., Hemmer-Hansen, J., Tomkiewicz, J., Bekkevold, D.,2009. Genetic population structure of European sprat Sprattus sprattus:differentiation across a steep environmental gradient in a small pelagic fish.Marine Ecology Progress Series 379, 213–224.

MacArthur, R.H., Wilson, E.O., 1967. The Theory of Island Biogeography. PrincetonUniv. Press, Princeton, NJ.

MacKenzie, B.R., Köster, F.W., 2004. Fish production and climate: sprat in the BalticSea. Ecology 85, 784–794.

Meskendahl, L., Herrmann, J.P., Temming, A., 2010. Effects of temperature and bodymass on metabolic rates of sprat, Sprattus sprattus L. Marine Biology 157, 1917–1927.

Möllmann, C., Köster, F.W., 2002. Population dynamics of calanoid copepods and theimplications of their predation by clupeid fish in the Central Baltic Sea. Journalof Plankton Research 24, 959–977.

Möllmann, C., Kornilovs, G., Fetter, M., Köster, F.W., 2004. Feeding ecology of centralBaltic Sea herring and sprat. Journal of Fish Biology 65, 1563–1581.

Möllmann, C., Diekmann, R., Müller-Karulis, B., Kornilovs, G., Plikshs, M., Axe, P.,2009. Reorganization of a large marine ecosystem due to atmospheric andanthropogenic pressure: a discontinuous regime shift in the Central Baltic Sea.Global Change Biology 15, 1377–1393.

Munk, P., 1993. Differential growth of larval sprat Sprattus sprattus across a tidalfront in the eastern North Sea. Marine Ecology Progress Series 99, 17–27.

Muus, B.J., Nielsen, J.G., 1999. Seafish. Blackwell Science, Oxford, UK, pp. 1–340.Nissling, A., 2004. Effects of temperature on egg and larval survival of cod (Gadus

morhua) and sprat (Sprattus sprattus) in the Baltic Sea – implications for stockdevelopment. Hydrobiologia 514, 115–123.

Osse, J.W.M., van den Boogaart, J.G.M., 1995. Fish larvae, development, allometricgrowth and the aquatic environment. ICES Marine Science Symposium 201, 21–34.

Overton, J.L., Meyer, S., Støttrup, J.G., Peck, M.A., 2010. The role of heterotrophicprotists in first feeding by cod (Gadus morhua L.) larvae. Marine EcologyProgress Series 410, 197–204.

Peck, M.A., Daewel, U., 2007. Physiologically-based limits to larval fish foodconsumption and growth: implications for individual-based models. MarineEcology Progress Series 347, 171–183.

Peck, M.A., Hufnagl, M., 2012. Can IBMs explain why most larvae die in the sea?Model scenarios and sensitivity analyses reveal research needs. Journal ofMarine Systems 93, 77–93.

Peck, M.A., Clemmesen, C., Baumann, H., Herrmann, J.-P., Stäcker, S., Temming, A.,2004. The growth-feeding relationship in post-larval sprat (Sprattus sprattus):comparison of somatic, nucleic acid- and otolith-based growth rates.International Council for the Exploration of the Sea. ICES CM 2004/L:25.

Peck, M.A., Clemmesen, C., Herrmann, J.-P., 2005. Ontogenic changes in theallometric scaling of the mass and length relationship in Sprattus sprattus.Journal of Fish Biology 66, 882–887.

Peck, M.A., Kühn, W., Clemmesen, C., Hinrichsen, H.-H., Holtappels, M., Huwer, B.,Pohlmann, T., 2007. Variability in growth rates of larval fish at frontal stations inthe southern North Sea: is the ‘‘mean’’ all that matters? GLOBEC InternationalNewsletter 14 (1), 70–71.

Pedersen, B.H., 1984. The intestinal evacuation rates of larval herring (Clupeaharengus L.) predating on wild plankton. Dana 3, 21–30.

Pepin, P., 1995. An analysis of the length-weight relationship of larval fish:limitations of the general allometric model. US Fisheries Bulletin 93, 419–426.

Pepin, P., Dower, J.F., 2007. Variability in the trophic position of larval fish in acoastal pelagic ecosystem based on stable isotope analysis. Journal of PlanktonResearch 29 (8), 727–737.

Pepin, P., Penney, R.W., 1997. Patterns of prey size and taxonomic composition inlarval fish: are there general size-dependent models? Journal of Fish Biology 51,84–100.

Pepin, P., Penney, R.W., 2000. Feeding by a larval fish community: impact onzooplankton. Marine Ecology Progress Series 204, 199–212.

Petereit, C., 2008. Influence of temperature and salinity on sprat (Sprattus sprattus)eggs and yolk sac larvae from contrasting environments. PhD Dissertation.University of Kiel, Kiel, Germany, 113 pp.

Petereit, C., Haslob, H., Kraus, G., Clemmesen, C., 2008. The influence of temperatureon the development of Baltic Sea sprat (Sprattus sprattus) eggs and yolk saclarvae. Marine Biology 154, 295–306.

Petitgas, P., Alheit, J., Peck, M.A., Raab, K., Irigoien, X., Huret, M., van der Kooij, J.,Pohlmann, T., Wagner, C., Zarraonaindia, I., Dickey-Collas, M., 2012. Anchovyexpansion in the North Sea. Marine Ecology Progress Series 444, 1–13.

Petrova, E.G., 1960. On the fecundity and maturation of Baltic sprat. TrudyVsesojuznyj naucno issledovatelskij. Institut Morskogo Rybnogo Xozjajstva iOkeanografii 42, 99–108 (in Russian).

Pörtner, H.O., Farrell, A.P., 2008. Physiology and climate change. Science 322, 690–692.

Pörtner, H.O., Peck, M.A., 2010. Climate change impacts on fish and fisheries:towards a cause and effect understanding. Journal of Fish Biology 77, 1745–1779.

Raab, K., Nagelkerke, L.A.J., Boerée, C., Rijnsdorp, A.D., Temming, A., Dickey-Collas,M., 2011. Anchovy Engraulis encrasicolus diet in the North and Baltic Seas.Journal of Sea Research 65, 131–140.

Rè, P., Gonçalves, E., 1993. Growth of sprat Sprattus sprattus larvae in the GermanBight (North Sea) as inferred by otolith microstructure. Marine Ecology ProgressSeries 96, 139–145.

Rijnsdorp, A., Peck, M.A., Engelhard, G.H., Möllmann, C., Pinnegar, J.K., 2009.Resolving the effect of climate change on fish populations. ICES Journal ofMarine Science 66, 1570–1583.

Rose, K.A., 24 co-authors, 2010. End-to-end models for the analysis of marineecosystems: challenges, issues, and next steps. Marine and Coastal Fisheries:Dynamics, Management, and Ecosystem Science 2, 115–130.

Rudstam, L.G., Hansson, S., Johansson, S., Larsson, U., 1992. Dynamics of planktivoryin a coastal area of the northern Baltic Sea. Marine Ecology Progress Series 80,159–173.

Safran, P., 1992. Theoretical analysis of the weight-length relationship in fishjuveniles. Marine Biology 112, 545–551.

Schultz, E.T., Conover, D.O., Ehtisham, E., 1998. The dead of winter: size dependentvariation and genetic differences in seasonal mortality among Atlanticsilverside (Atherinidae: Menidia menidia) from different latitudes. CanadianJournal of Fisheries and Aquatic Science 55, 1149–1157.

Schulz, J., Barz, K., Peck, M.A., Schmidt, J.O., Hansen, F., Peters, J., Renz, J., Dickmann,M., Mohrholz, V., Dutz, J., Hirche, H.-H., submitted for publication. Spatial andtemporal habitat partitioning by zooplankton in the Bornholm Sea (centralBaltic Sea). Progress in Oceanography.

Shields, R.J., 1989. Studies of growth and nutritional status in 0-group sprat, Sprattussprattus (Clupeidae), using otolith microstructure and lipid analyticaltechniques. PhD Thesis. University of Wales, Bangor.

Shvetsov, F.G., Starodub, M.L., Sidrevits, L.L., 1983. The daily feeding rhythm of theBaltic sprat, Sprattus sprattus balticus (Clupeidae). Journal of Ichthyology 23 (2),99–105.

Sirotenko, M.D., Sorokalit, L.K., 1979. Seasonal changes in the food of theMediterranean sprat, Sprattus sprattus phalericus. Journal of Ichthyology 19(5), 37–51.

Smith, R.L., Paul, A.J., 1990. Seasonal changes in energy and energy cost of spawningin Gulf of Alaska Pacific cod. Journal of Fish Biology 36, 307–316.

Stepputtis, D., 2006. Distribution patterns of Baltic sprat (Sprattus sprattus L.) –causes and consequences. PhD Thesis. Leibniz Institute for Marine Research,Kiel, 198 pp.

Szypula, J., 1985. Comparative studies on herring and sprat feeding in the southernBaltic within 1978–1982. Acta Ichthyologica et Piscatoria 15 (2), 75–89.

Szypula, J., Grygiel, W., Wyszynski, M., 1997. Feeding of Baltic herring and sprat inthe period 1986–1996 in relation to their state and biomass. Bulletin of the SeaFisheries Institute 3 (142), 73–83.

Thompson, B.M., Milligan, S.P., Nichols, J.H., 1981. The development rates of Sprat(Sprattus sprattus L.) eggs over a range of temperatures. International Council forthe Exploration of the Sea. ICES CM 1981/H:15.

Valenzuela, G.S., Vargas, C.A., 2002. Comparative growth rate of Sprattus sprattus inrelation to physical and oceanographic features in the North Sea. Archive ofFisheries and Marine Research 49 (3), 213–230.

Vieira, V.L.A., Johnston, I.A., 1992. Influence of temperature on muscle-fibredevelopment in larvae of the herring Clupea harengus. Marine Biology 112,333–341.

Voss, R., Köster, F.W., Dickmann, M., 2003. Comparing the feeding habits of co-occurring sprat (Sprattus sprattus) and cod (Gadus morhua) larvae in theBornholm Basin, Baltic Sea. Fisheries Research 63, 97–111.

Voss, R., Clemmesen, C., Baumann, H., Hinrichsen, H.-H., 2006. Baltic sprat larvae:coupling food availability, larval condition and survival. Marine EcologyProgress Series 308, 243–254.

Voss, R., Schmidt, J.O., Schnack, D., 2007. Vertical distribution of Baltic sprat larvae:changes in patterns of diel migration? ICES Journal of Marine Science 64 (5),956–962.

Voss, R., Dickmann, M., Schmidt, J.O., 2009. Feeding ecology of sprat (Sprattussprattus L.) and sardine (Sardina pilchardus W.) larvae in the German Bight,North Sea. Oceanologia 51, 117–138.

Voss, R., Peck, M.A., Hinrichsen, H.-H., Clemmesen, C., Baumann, H., Stepputtis, D.,Bernreuther, M., Schmidt, J.O., Temming, A., Köster, F.W., submitted forpublication. Recruitment processes in Baltic sprat – a re-evaluation ofGLOBEC-Germany hypotheses. Progress in Oceanography.

56 M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57

Author's personal copy

Wahl, E., Alheit, J., 1988. Changes in the distribution and abundance of sprat eggsduring spawning season. International Council for the Exploration of the Sea.ICES CM 1988/H:45.

Whitehead, P.J.P., 1985. Clupeoid fishes of the world. An annotated and illustratedcatalogue of the herrings, sardines, pilchards, sprats, shads, anchovies and wolf-herrings. Part 1 Chirocentridae, Clupeidae and Pristigasteridae. FAO FisheriesSynopsis 125 (7), 1–303.

Wieland, K., Zuzarte, F, 1991. Vertical distribution of cod and sprat eggs and larvaein the Bornholm Basin (Baltic Sea) in 1987–1990. International Council for theExploration of the Sea. ICES CM 1991/J:37.

Wosnitza, C., 1974. Die Nahrung von Fischbrut in der westlichen Ostsee.Berichte der Deutschen Wissenschaftlichen Kommission für Meeresforschung24, 79–92.

M.A. Peck et al. / Progress in Oceanography 103 (2012) 42–57 57