Population ecology and prey consumption by fathead minnows in

Population ecology and prey consumption byfathead minnows in prairie wetlands: importanceof detritus and larval fish

Introduction

Fathead minnow Pimephales promelas is a commonresident of semipermanent and permanently floodedwetlands throughout much of the prairie potholeregion (PPR) in North America (Stewart & Kantrud1971; Peterka 1989). Aquatic invertebrates are alsoabundant in these habitats, and comprise an importantfood resource for vertebrate consumers, including fish,birds and amphibians (Euliss et al. 1999). However,recently, fathead minnow populations have beenshown to strongly influence community characteristicsin Minnesota’s PPR wetlands, including reducinginvertebrate populations, and creating conditions thatfavour high phytoplankton biomass, low water trans-parency and reduced submerged aquatic vegetation(Zimmer et al. 2000, 2001a, 2003).

Natural history characteristics (omnivory – Held &Peterka 1974; Price et al. 1991; high tolerance tohypoxic conditions – Klinger et al. 1982; rapid growth– Held & Peterka 1974; high recruitment potential –Payer & Scalet 1978) make the fathead minnow welladapted to shallow wetland habitats, and contribute totheir widespread occurrence in PPR wetlands. Thefathead minnow is also a fractional spawner, withpopulations producing multiple cohorts of larvaethroughout the summer once water temperatures reach18 �C (Dobie et al. 1956). These fish grow quicklyand can reach sexual maturity and spawn at the end oftheir first summer (Markus 1934; Held & Peterka1974). Zooplankton, macroinvertebrates, and detritushave all been found to be important components offathead minnow diets at various times of the year(Held & Peterka 1974; Price et al. 1991; Zimmer et al.

Ecology of Freshwater Fish 2007: 16: 282–294Printed in Singapore Æ All rights reserved

� 2007 State of Minnesota, Department of Natural ResourcesJournal compilation � 2007 Blackwell Munksgaard

ECOLOGY OFFRESHWATER FISH

282 doi: 10.1111/j.1600-0633.2006.00220.x

Herwig BR, Zimmer KD. Population ecology and prey consumption byfathead minnows in prairie wetlands: importance of detritus and larval fish.Ecology of Freshwater Fish 2007: 16: 282–294. � 2007 State ofMinnesota, Department of Natural Resources. Journal compilation� 2007 Blackwell Munksgaard

Abstract – The fathead minnow Pimephales promelas occurs in highdensities in wetlands of the prairie pothole region (PPR) of North America,but food resources sustaining these populations are poorly known. Weassessed population dynamics and prey consumption of fathead minnowpopulations in three PPR wetlands for 2 years. Fish density peaked at 107fish per m2 for all age classes combined. Larval and juvenile fishdominated these populations in terms of abundance and accounted for83% of total prey consumption. Detritus dominated fish diets, representing53%, 40% and 79% of diet mass for larval, juvenile and adult fishrespectively. Detritus consumption was positively related to minnowdensity and negatively related to invertebrate abundance, but only foradult fish. Seasonal production:biomass ratios were unrelated toproportions of detritus in the diet for all ages of fish, indicating that detritusis an important food resource capable of meeting metabolic demandsand sustaining fish growth in high-density populations. Detritusconsumption may also weaken links between abundance of invertebrateprey and minnows, promoting dense fish populations with strong,consistent influences on wetland ecosystems.

B. R. Herwig1, K. D. Zimmer2

1Division of Fish and Wildlife – FisheriesResearch, Minnesota Department of NaturalResources, Bemidji, 2Department of Biology,University of St Thomas, St Paul, MN, USA

Key words: fathead minnow; Pimephales prom-elas; bioenergetics models; consumption esti-mates; fish production; larval fish; detritus

Correspondence: Brian R. Herwig, Division ofFish and Wildlife – Fisheries Research, MinnesotaDepartment of Natural Resources, 2114 BemidjiAvenue, Bemidji, MN 56601, USA;e-mail: [email protected]

Accepted for publication November 18, 2006

2006). These characteristics all contribute to highfathead minnow densities and biomass in prairiewetlands, as well as high consumption rates of aquaticinvertebrates (Duffy 1998). However, to date, researchon prey consumption by the fathead minnow haslargely ignored consumption by larval fish. Further-more, the overall importance of detritus as a foodresource for all ages of minnows is poorly understood.Patterns of metabolic allometry for fish show that

mass-specific consumption rates are inversely relatedto fish size (Post 1990). Thus, gram for gram, larvalfish have the potential to exert greater impacts on preypopulations and energy flow in wetland ecosystemsthan do older, large-bodied adults. Implications of thisrelationship are particularly important in PPR wetlandsas young-of-the-year (YOY) and larval fish oftendominate fathead minnow populations in these sys-tems. Payer & Scalet (1978) estimated 99% of theannual production of a fathead minnow populationwas contributed by YOY fish, and estimated that 194adult fish resulted in over 126,000 recruits in just3 months.The role of detritus as a food resource for fathead

minnows is poorly known. Consumption of detritusweakens the link between the abundance of inverteb-rate prey and abundance of fish by allowing higherpopulation-level survival during times of lean inver-tebrate abundance, resulting in higher, more consistentpredation pressure on invertebrates relative to fishpopulations feeding solely on invertebrates (Polis &Strong 1996; Schaus & Vanni 2000; Schindler &Scheuerell 2002). Additionally, consumption of detri-tus and subsequent excretion translocates nutrients andrepresents a ‘new’ source of nutrients for phytoplank-ton, potentially increasing their overall abundance(Schaus & Vanni 2000; Zimmer et al. 2006). Thus,influences of fathead minnow populations on wetlandecosystems may depend, in part, on the relativeimportance of detritus in the diet. Consumption ofdetritus may also influence dynamics of minnowpopulations (such as mass-specific production rates ofthe fish), as detritus has lower concentrations of bothenergy and protein relative to invertebrate prey(Bowen et al. 1995). Consequently, detritus has usu-ally been considered a supplemental food source thatallows the fathead minnow to survive when other morenutritional food sources are low (Held & Peterka 1974;Price et al. 1991). More recently, it has been suggestedthat consumption of lower quality, readily availabledetritus, combined with higher quality, difficult-to-catch invertebrate prey may serve as a mechanismfor rapid growth of fathead minnows (Lemke &Bowen 1998).Here, we describe population characteristics, and

estimate consumption of invertebrate and detrital foodresources in adult, juvenile and larval fathead minnow

cohorts in three prairie wetlands. Our goals were to:(1) estimate and contrast consumption of detritus andinvertebrate prey among larval, juvenile, and adultfish; (2) estimate rates of biomass and energy flowthrough the fathead minnow populations; (3) testwhether consumption of detritus was related toabundance of invertebrate prey; and (4) test whethermass-specific production rates of minnow populationswere related to the amount of detritus in fish diets.

Methods

We studied fathead minnow populations in threewetlands in western Minnesota. Study wetlandsranged in size from 6Æ4 to 13Æ0 ha, and maximumwetland depths were 1Æ9–2Æ6 m (Table 1).

Fathead minnow population characteristics and growth

Density estimates of juvenile (20–40 mm TL) andadult (>40 mm TL) fathead minnows were obtainedusing square 1-m2 pop nets (Dewey et al. 1989) fittedwith 0Æ5 mm (bar measure) mesh. Pop nets consist of abuoyant top frame (PVC pipe and foam insulation)fitted with 1 m2 netting attached to four sides and thebottom, but open on top. The net is then temporarilysecured to a weighted bottom frame and deployed.After a waiting period, the device is triggered remotelysending the top frame and netting rapidly up throughthe water column, capturing small fishes and thusproviding a quantitative estimate of minnow density inthe selected area sampled. Twelve samples werecollected using pop nets that were randomly distri-buted in the 0Æ1- to 1Æ0-m depth zone of each wetlandon each sampling date. Populations were sampledonce every 2 weeks in 2001 and once every 3 weeksin 2002 from mid-May to mid-September each year.Pop nets were allowed to sit undisturbed on thewetland sediment for 1 h prior to being triggered.Although our sampling was restricted to 0Æ1- to 1-mwater depths, our previous work with pop nets hasindicated densities of larval, juvenile and adult min-nows do not differ with depth in these systems (B.R.Herwig and K.D. Zimmer, unpublished data). Min-nows captured in the pop nets were counted, and up to40 fish per sample were measured for total length to

Table 1. Classification, surface area, maximum depth and other fish speciespresent for the three study sites located in west-central Minnesota.

Wetland Classification Area (ha)Maxdepth (m)

Other fishspecies present

Stammer 4 9Æ1 1Æ88 Culaea inconstansFroland 5 6Æ4 2Æ08 NoneBellevue 5 13Æ0 2Æ55 Culaea inconstans

Classification of Stewart & Kantrud (1971).

Fathead minnows in prairie wetlands

283

the nearest 1 mm. Larval (<20 mm) fathead minnowdensity estimates were obtained using a larvalichthyoplankton push net (0Æ5 m diameter, 0Æ8 mmmesh) at three fixed transects running parallel to shore,thus providing a spatially integrated estimate ofabundance in each wetland. The water depth alongeach transect was approximately 1 m, and wecollected each sample by pushing the ichthyoplanktonnet alongside the bow of the boat at approximately5 km h)1 just below the surface for 120 s. A calib-rated flow meter mounted inside the net was used todetermine the volume of water sampled for larval fishdensity estimates. The sampling schedule for larvalminnows was identical to that used for samplingjuveniles and adults. Larval minnows captured in theicthyoplankton samples were preserved in 95% eth-anol in the field, and were later counted and measured(TL to the nearest 1 mm; up to 100 fish per date). Weestimated the wetland-wide population size of adultsand juveniles by multiplying the density estimatesfrom pop netting by wetland surface area. Larvalpopulation size was estimated by multiplying larvaldensities from ichthyoplankton tows by wetlandvolume. Depth profiles of each wetland were mappedwith a GPS, and wetland surface area and volumewere estimated using Surfer software (Golden Soft-ware 1997).

The fathead minnow spawns fractionally and iscapable of producing multiple cohorts each year;therefore, standard aging methods cannot be used toidentify cohorts in fathead minnow populations (Duffy1998). Thus, we used the length–frequency distribu-tion estimated from our population sampling toidentify and track the growth of individual cohortsthroughout the sampling season (Duffy 1998).Length–frequency histograms were constructed foreach sampling date and modal lengths were identifiedbased on visual inspection (see Fig. 1 for an example).Fish whose sizes were within ±20% of a mode weregrouped as a cohort (Duffy 1998; Fig. 1). Meanlengths of cohorts were converted to mean weightsbased on weight–length regressions developed forfathead minnow populations across the three studywetlands (B.R. Herwig, unpublished data), and growthof individuals in each cohort was measured as theincrease in average weight between one sampling dateand the next.

Fathead minnow diets

Diets of adult, juvenile and larval fathead minnowwere assessed in June, July and August of each year.Ten fish were collected from each size class in eachwetland on each date. Larvae were collected using anichthyoplantkon net, while juveniles and adults werecollected using a beach seine. Fish were anaesthetised

with a lethal dose of MS-222TM, preserved in a 10%formalin solution, and later transferred to 95% ethanol.Only the anterior one-third of the intestinal tract wasdissected and analysed because items lower in theintestinal tract are severely digested and difficult toidentify (Duffy 1998). We identified diet items to thelowest feasible taxonomic level (typically genus forzooplankton, and either family or order for macroin-vertebrates) and, where possible, measured bodylengths of individual items. Lengths were thenconverted to weights using length–weight regressions(Smock 1980; McCauley 1984, B.R. Herwig & K.D.Zimmer, unpublished data). After all invertebrateswere removed from the sample the remaining detrituswas filtered onto a preweighed 0Æ7 lm glass–fibrefilter to obtain both wet and dry weights. To obtain anaccurate wet weight, excess water was removed byplacing the filter on a perforated funnel, and a lightvacuum was applied using an electric pump. Filterswere then placed in a drying oven at 60 �C for 24 h toobtain dry weights.

Bioenergetics modelling

We estimated prey consumption for fathead minnowpopulations using Fish Bioenergetics 3.0 (Hansonet al. 1997). Modelling consumption using the bioen-ergetics model requires constructing energy budgetsvia the mass balance equation C ¼ G + R + F + Uwhere C is the energy ingested, G the observedgrowth, R the energy used in metabolism, F theegestion and U the excretion (Hewett & Johnson1987). We used physiological parameters developedby Duffy (1998) for our adult and juvenile fatheadminnow cohorts, but these adult-derived parametersdid not fit larval fish metabolism due to metabolic

Length (mm)10 20 30 40 50 60 70

Num

ber

0

3

6

9

12

15

18

Cohort 1

Cohort 2

Cohort 3

Fig. 1. Length frequency distribution of fathead minnow collectedfrom Bellevue wetland on 22 July 2002 to illustrate the method ofcohort separation. Horizontal lines indicate ±20% of the modalvalue for each cohort.

Herwig & Zimmer

284

allometry (Post 1990). We used bioenergetics param-eters developed by Post (1990) for larval yellow perchPerca flavescens to model our larval fathead minnowcohorts, the most similar freshwater species for whichlarval modelling parameters have been developed. Thecomplete lists of bioenergetics parameters used tomodel juvenile and adult minnows, and larvae, can befound in Duffy (1998) and Post (1990) respectively.We modelled individual cohorts, and then compiledestimates for larval, juvenile and adult fish. Wemodelled all three age groups from 15 May to 30August in 2001 and from 15 May to 2 September in2002.Field data required to parameterise the model

include fathead minnow diet composition, populationsize, growth and mortality rates, predator and preyenergy densities, and water temperature. Diet compo-sition was modelled as the proportion of wet weight ofeach diet item consumed by each life stage on each ofthe three sampling dates. Diets were assumed toremain constant between sampling periods. When acohort grew into a subsequent life stage, the diet dataappropriate to that life stage was applied. Populationsize for each cohort was specified as the proportion ofthe total population represented by that cohort. Eachcohort was then tracked from one sampling period tothe next, providing a measure of fish growth (changein mass), production (accumulation of biomass) andmortality (loss of biomass). Energy density for thefathead minnow was taken from Duffy (1998), energydensity for invertebrates from Cummins & Wuychek(1971), Schindler et al. (1971), Driver et al. (1974)and Vijverberg & Frank (1976), and energy in detritusfrom Penczak (1985). Water temperatures wererecorded hourly in each wetland using temperatureloggers suspended 1 m below the surface. Mean dailywater temperatures were entered into the model foreach wetland for each model day. We modelledspawning losses as a discrete event, where femaleslost 11% body mass upon reaching 55 mm TL. Thiswas based on observations made by Duffy (1998),who found that 55 mm was the most common lengthfor gravid females and that on average females lost11% of their mass following spawning.All calculations were performed on a daily interval.

We combined the model outputs for each cohort toestimate prey consumption by larval, juvenile, adultand the entire fathead minnow population in eachwetland on each day. Model output was also used todetermine gross and net production, production:bio-mass ratios, biomass turnover time (TT) (calculated asbiomass:gross production), and mortality rates of eachage class.Relative abundance of aquatic invertebrates in each

wetland was measured using vertically oriented activ-ity traps (ATs) (Murkin et al. 1983) that were deployed

at the same time we assessed fathead minnow diets.Ten ATs were set 30 cm below the water in eachwetland, retrieved after 24 h, and contents concentra-ted by passage through a 140-lm mesh. We thendetermined the total number of invertebrates capturedin the ATs (summed across taxa) that were alsoobserved in diets of the fish. The total number ofinvertebrates for each wetland–date (N ¼ 18) wassubsequently used to reflect relative abundance ofinvertebrate prey on each date we sampled fish diets.

We used simple linear regression to test whether fishdensity or abundance of aquatic invertebrates influ-enced consumption of detritus by larval, juvenile andadult fathead minnow populations across samplingdates. Consumption of detritus was expressed as theproportion of total consumption for each age class onthe 18 wetland–date combinations, and was arcsinetransformed to meet the assumptions of linear regres-sion. Fish density in each age class and the totalnumber of aquatic invertebrates captured in ATs oneach wetland–date were both log-transformed tohomogenise variance. We then regressed log inverteb-rate abundance for each wetland–date combinationagainst arcsine-transformed proportion of detritus inthe diet on the same date. Similarly, we regressed logfish abundance against arcsine-transformed proportionof detritus in the diet on the same date. This analysisused six samples from each wetland, and autocorre-lation among the six samples may cause biasedestimates and significance tests in the regressionanalysis using all 18 samples (Bence 1995). Thus,we used the Durbin–Watson test for first-order auto-correlation to assess the degree of autocorrelation forlarval, juvenile and adult fish samples within eachwetland (Bence 1995). This analysis was done for boththe regression of log fish density on arcsine proportionof detritus and the log invertebrate abundance onarcsine proportion of detritus. We detected no signi-ficant autocorrelation among any samples (allP > 0Æ05), and the overall autocorrelation wasweak among samples (average Durbin–WatsonP-value ¼ 0Æ40). However, autocorrelation is difficultto detect with sample sizes as small as ours, so thesubsequent regression analyses should still be consid-ered approximate.

We also assessed the influence of detritus consump-tion on production rates for each of our three ageclasses by regressing the log-transformed, seasonalfish P:B ratio in each wetland in each year on thearcsine-transformed, proportion of diet represented bydetritus.

Results

Larval and juvenile fathead minnow densities inStammer and Bellevue were substantially higher in


285

2001 than in 2002 (Fig. 2). Larval and juvenile fishdominated in these wetlands in 2001, with adultdensities seldom exceeding two fish per m2. In 2002,juvenile and adult fish dominated in Stammer andBellevue. Population trends differed considerably inFroland, where densities of all size classes remainedlow throughout 2001 and during early 2002, butdensities of all size classes increased during the laterpart of 2002.

Larval and juvenile densities fluctuated dramaticallyin all sites during 2001 and 2002. By early August2001, summed fish density exceeded 65 fish per m2 inboth Stammer and Bellevue (Fig. 2). Peak biomasswas observed on the same dates as peak density, with156 kg ha)1 in Bellevue and 63 kg ha)1 in Stammer.

During 2002, juvenile and adult densities reached acombined maximum of 17 fish per m2 in Froland, 30fish per m2 in Stammer (both in early September) and34Æ5 fish per m2 in Bellevue in mid-July (Fig. 2).Biomass peaked in mid-July in Bellevue(167 kg ha)1), in mid-May in Stammer (252 kg ha)1),and in both mid-May and early September in Froland(60 kg ha)1).

The fathead minnow populations within each wet-land produced four to seven larval cohorts each year,but recruitment to the juvenile stage was highlyvariable. At least some recruitment to juvenile wasobserved for most wetland-years, except for Froland in2001 where all seven cohorts failed to recruit tojuvenile sizes. Growth was variable among wetlands,but we generally observed fast growth, with some fishreaching 2 g in their first growing season. Few adultfish exceeded 4 g, with the exception of one cohortfrom Stammer in early 2002, which resulted in highstanding stock biomass and prey consumption on thisdate (Fig. 3).

Average standing stock biomass of fathead minnowpopulations summed across age classes ranged from 5to 76 kg ha)1 (Table 2). Average biomass was high inStammer in both years (40 and 76 kg ha)1) and inBellevue in 2001 (59 kg ha)1), but was low in Frolandin 2001 (5 kg ha)1). Gross production (GP) summedacross age classes in 2001 was highest in Stammer andBellevue (3Æ12 and 2Æ97 kg ha)1 day)1 respectively),but was only 0Æ04 kg ha)1 day)1 in Froland (Table 2).In 2002, total GP was similar among all wetlands andranged from 0Æ93 to 1Æ43 kg ha)1 day)1. Gross pro-duction was dominated by larval and juvenile fish.Averaged across all wetland-years, larval and juvenilefish accounted for 41% and 37% of total GP,respectively, while representing just 19% (larval) and28% (juvenile) of the average biomass. Mortality rateswere generally high for all age classes in all wetlands,and rates across age classes ranged from 5% to 10%per day (Table 2). High production rates coupled withrelatively small standing stock biomass resulted inrapid TT in all age classes, and was highest in larvalfish, ranging from 7 to 29 days (mean ¼ 14 days)(Table 2). Scaled allometrically, TT ranged from 15 to28 days in juvenile fish (mean ¼ 20 days) and from41 to 163 days in adults (mean ¼ 72 days).

Total daily consumption rates summed across ageclasses ranged from 0Æ01 to 38Æ91 kg ha)1 day)1

(Table 3). Total prey consumption varied greatlybetween years in all three wetlands, reflecting sub-stantial differences in fish abundance (Fig. 2). Inter-annual differences were most pronounced in Froland,where prey consumption increased approximately28-fold from 2001 to 2002, concurrent with increasesin densities of all fathead minnow size classes during2002 (Table 3, Fig. 2). Larval fish accounted for the

Date

0

5

10

15

20

25

30

0

20

40

60

80

100

120

140

StammerFrolandBellevue

wonnim daehta

F#·

m–2

)( ytisned

0

10

20

30

40

50

60

70

Larval(a)

(b)

(c)

Juvenile

Adult

29 May

2001

30 Aug 20 May 3 Sep

29 May 30 Aug 20 May 3 Sep

29 May 30 Aug 20 May 3 Sep

2002

Fig. 2. Density estimates for (a) larval (no. of fish <20 mm m)2),(b) juvenile (no. of fish 20–40 mm m)2) and (c) adult (no. of fish>40 mm m)2) fathead minnow populations in Stammer (filledcircles), Froland (open circles) and Bellevue (filled triangles)wetlands throughout summer 2001 and 2002 (±1 SE).

Herwig & Zimmer

286

majority (69%) of total food consumption by fatheadminnows, despite representing only 19% of the totalbiomass in the population.Fathead minnow diets were highly variable and

included diverse prey items, including 31 differenttaxonomic items in total, plus detritus. It is notable thatall size classes of minnows consumed each of themajor food categories (rotifers, zooplankton, macro-invertebrates and detritus) in their diet at some pointduring the study. Detritus was an important foodresource for all fish sizes, and averaged across

wetland-years it accounted for 53%, 40% and 79%of diet by mass for larval, juvenile and adult fishrespectively. Overall, detritus accounted for 55% oftotal fish consumption (Table 3). Larval fish accountedfor 66% of the total mass of consumed detritus, despitethe higher selectivity for detritus by adult fish and thehigher biomass of adult fish. In contrast to theubiquitous reliance on detritus, consumption of inver-tebrate prey differed substantially among fish sizeclasses. In addition to detritus, larval minnowsprimarily consumed non-rotifer zooplankton (21% of

0

5

10

15

20

25

Stammer

0

20

40

60

80

100

120

140

160

(a)

(b)

(c)

Total preyDetritusZooplanktonMacroinvertebrates

Larval

0

10

20

30

40

50

Juvenile

0

1

2

3

4

5g·ha

–1)

k( noitpmusnoc yerp deta

mitsE

0

2

4

6

8

10

Adult

0

10

20

30

40

50

60

29 May 20 May 3 Sep2001

Date

30 Aug2002

Fig. 3. Estimated prey consumption (kgha)1 day)1) from a bioenergetics modelapplied to (a) larval, (b) juvenile and (c) adultfathead minnow size classes in Stammerwetland in 2001 (left panel) and 2002 (rightpanel). Data points shown are for those dateswhen population size was directly estimatedin the field.

Table 2. Mean biomass (B), gross production(GP), turnover time (TT, biomass/gross produc-tion) and natural mortality (M) for larval, juvenileand adult fathead minnow populations from threeMinnesota study wetlands in 2001 and 2002.

Wetland

Stammer Froland Bellevue

Larval Juvenile Adult Larval Juvenile Adult Larval Juvenile Adult

2001B (kgÆha)1) 15Æ62 15Æ57 9Æ19 0Æ20 0Æ04 4Æ89 17Æ84 20Æ44 21Æ01GP (kgÆha)1Æday)1) 1Æ87 1Æ04 0Æ21 0Æ03 0Æ01 0Æ03 1Æ30 1Æ16 0Æ51M (kgÆha)1Æday)1) 1Æ51 1Æ29 0Æ97 0Æ01 0Æ03 0Æ34 0Æ87 0Æ94 1Æ95TT (days) 8Æ35 14Æ97 43Æ76 6Æ67 20Æ00 163Æ00 13Æ72 17Æ62 41Æ202002B (kgÆha)1) 3Æ61 5Æ51 67Æ15 5Æ71 7Æ10 13Æ28 1Æ76 16Æ43 8Æ31GP (kgÆha)1Æday)1) 0Æ17 0Æ20 1Æ06 0Æ42 0Æ32 0Æ19 0Æ10 0Æ80 0Æ12M (kgÆha)1Æday)1) 0Æ01 0Æ14 3Æ66 0Æ33 0Æ02 0Æ95 0Æ02 1Æ33 1Æ31TT (days) 21Æ24 27Æ55 63Æ35 17Æ84 22Æ19 69Æ89 17Æ60 20Æ54 69Æ25


287

total consumption) and rotifers (20%), while juvenilefish consumed macroinvertebrates (48%) andzooplankton (12%) and adult fish consumed mostlymacroinvertebrates (19%) (Figs 3–5).

Prey consumption rates reflected population sizes,with peak periods corresponding to maximum fatheadminnow densities (Figs 2–5). Abundance of fish in thethree size classes fluctuated due to timing of repro-duction, such that adult and juvenile consumptionrates generally peaked in June to early July, whilemaximum consumption by larval fish peaked in mid-to late August. Comparing the timing of maximum fishdensities and diet composition of the three size classesindicates that consumption of macroinvertebratespeaked in June, zooplankton and rotifers in late Julyto early August, and detritus in early to late August.

Relationships among consumption of detritus, fishdensity and abundance of aquatic invertebrates werevariable, but more robust with increasing fish size(Fig. 6). We detected no relationship between abun-dance of larval fish and the proportion of detritus intheir diet (F1,16 ¼ 3Æ81, P ¼ 0Æ068, r2 ¼ 0Æ19), norany relationship between invertebrate abundance andconsumption of detritus (F1,16 < 0Æ01, P ¼ 0Æ98,r2 < 0Æ01). These relationships were stronger forjuvenile fish, with a significant negative relationshipobserved between detritus in the diet and abundance of

Table 3. Estimated cumulative consumption (kg ha)1 day)1) of major preytypes from mid-May to mid-September by larval, juvenile and adult fatheadminnow populations in three Minnesota study wetlands in 2001 and 2002.

Life stage

Major prey taxa

Zooplankton Rotifers Macroinvertebrates DetritusTotalprey

Stammer 2001Larval 0Æ83 12Æ96 0Æ25 15Æ00 29Æ04Juvenile 0Æ15 0 3Æ67 2Æ25 6Æ06Adult 0Æ16 0 0Æ62 0Æ54 1Æ32Stammer 2002Larval 1Æ95 0Æ92 1Æ36 1Æ54 5Æ77Juvenile 0Æ11 0 0Æ46 1Æ29 1Æ86Adult 0Æ07 0Æ01 0Æ87 15Æ77 16Æ73Froland 2001Larval 0Æ33 0 0 0 0Æ33Juvenile 0 0 0Æ01 0 0Æ01Adult 0Æ04 0 0Æ31 0Æ10 0Æ44Froland 2002Larval 2Æ71 1Æ05 0 13Æ34 17Æ11Juvenile 0Æ39 0Æ04 0Æ23 2Æ02 2Æ68Adult 0Æ02 0Æ01 0Æ81 0Æ76 1Æ59Bellevue 2001Larval 14Æ33 4Æ96 0 19Æ6 38Æ91Juvenile 1Æ08 0 2Æ17 0Æ75 4Æ00Adult 0Æ32 0 1Æ16 0Æ87 2Æ36Bellevue 2002Larval 0Æ49 0Æ17 2Æ32 0 2Æ97Juvenile 0Æ42 0 2Æ39 1Æ06 3Æ87Adult 0Æ01 0 0Æ54 0Æ28 0Æ79

0

10

20

30

40

50

Froland(a)

(b)

(c)

0

1

2

3

4

5Total preyDetritusZooplanktonMacroinvertebrates

Larval

0.00

0.02

0.04

0.06

0.08

0.10

0

4

8

12

16g·ha

–1)


mitsE

0

1

2

3

4

0

2

4

6

8

Juvenile

Adult

29 May2001

Date

30 Aug 20 May 3 Sep2002

Fig. 4. Estimated prey consumption(kg ha)1 day)1) from a bioenergetics modelapplied to (a) larval, (b) juvenile and (c)adult fathead minnow size classes inFroland wetland in 2001 (left panel) and2002 (right panel). Data points shown arefor those dates when population size wasdirectly estimated in the field.

Herwig & Zimmer

288

invertebrates (F1,16 ¼ 4Æ84, P ¼ 0Æ043, r2 ¼ 0Æ23) butno statistically significant relationship between pro-portion of detritus in the diet and fish density(F1,16 ¼ 4Æ21, P ¼ 0Æ056, r2 ¼ 0Æ21). Adult con-sumption of detritus was positively related to adultfish density (F1,16 ¼ 9Æ12, P ¼ 0Æ008, r2 ¼ 0Æ36) andadult consumption of detritus was negatively related toinvertebrate abundance (F1,16 ¼ 6Æ28, P ¼ 0Æ023,r2 ¼ 0Æ28) (Fig. 6).Detritus as a proportion of the total diet over the

sampling season varied dramatically among the sixwetland-years, ranging from 0% to 78% for larval fish,0–76% for juvenile fish and 23–94% for adult fish.However, consumption of detritus had little effect onfish production, as we detected no significant relation-ship between seasonal P:B ratios and proportion ofdetritus in the diet for adult (F1,4 ¼ 0Æ10, P ¼ 0Æ769,r2 ¼ 0Æ02), juvenile (F1,4 ¼ 0Æ57, P ¼ 0Æ492,r2 ¼ 0Æ02) or larval fish (F1,4 ¼ 0Æ21, P ¼ 0Æ668,r2 ¼ 0Æ05) (Fig. 7).

Discussion

Fathead minnow populations in our study exhibitedprolific recruitment (up to seven cohorts per summer),reached high standing stock biomass (up to

252 kg ha)1), and experienced high levels of naturalmortality. As a consequence, smaller bodied larval andjuvenile fish often dominated the minnow populationsin terms of numbers, biomass and production. Typic-ally, larval fish have been considered ‘too dilute’ tohave significant impacts on prey populations (Cushing1983), and are often ignored in models describing theroles of fish in shallow lakes. However, metabolicallometry generates an inverse relationship betweenmass-specific consumption rates and fish body size,causing influences of larval and juvenile fish on preypopulations to be potentially much higher comparedwith similar adult biomass (Post 1990). Larval andjuvenile production was appreciable in our study sites(78% of total GP), and accounted for 83% of totalconsumed prey, despite comprising only 47% of thetotal fish mass. To our knowledge, Duffy (1998)represents the only other attempt to estimate preyconsumption by fathead minnow populations, but hisconsumption estimates were restricted to fish classifiedas adults in our study (>40 mm TL). Across four SouthDakota wetlands, Duffy estimated average prey con-sumption rates of adult minnows ranging from 4Æ3 to16Æ7 kg ha)1 day)1 with a mean of 8Æ3 kg ha)1 day)1.In contrast, our average daily consumption ratessummed across larval, juvenile and adult fish ranged

0

2

4

6

8

0

20

40

60

80

100

120

140

Total PreyDetritusZooplanktonMacroinvertebrates

Larval

(a)

(b)

(c)

Bellevue

g·ha

–1)


mitsE

0

2

4

6

8

Juvenile

0

2

4

6

8

10

Adult

0

5

10

15

20

25

0

2

4

6

8

10

12

29 May 3 Sep2001

Date

30 Aug 20 May2002

Fig. 5. Estimated prey consumption(kg ha)1 day)1) from a bioenergetics modelapplied to (a) larval, (b) juvenile and (c)adult fathead minnow size classes in Belle-vue wetland in 2001 (left panel) and 2002(right panel). Data points shown are forthose dates when population size was dire-ctly estimated in the field.


289

from 0Æ8 to 36Æ4 kg ha)1 day)1, and the mean of22Æ6 kg ha)1 day)1 was 2Æ7-fold higher than reportedby Duffy (1998). Despite being largely ignored inprevious studies, our results highlight the importanceof larvae and juveniles when describing theeffects of fathead minnow populations on aquaticecosystems.

We observed a maximum density and biomass offathead minnow of 107 fish per m2 and 252 kg ha)1,respectively, which was generally lower than estimatesfor other prairie wetlands. Duffy (1998) reported thatstanding stock biomass of adult fathead minnows inthree South Dakota wetlands exceeded 300 kg ha)1 onone or more sampling dates, with a maximum biomassof 482 kg ha)1 and density of 43 fish per m2. Incontrast, Zimmer et al. (2001a) estimated populationsof all fish sizes and observed peaks of 369 kg ha)1

and densities of 570 fish per m2 in one Minnesotawetland. In aquatic systems outside of the prairiepothole region, fathead minnow density has beenreported to range from 1Æ8 fish per m2 in an Alberta

pond (Price et al. 1991) to 25 fish per m2 in Michiganponds (Spencer & King 1984). In culture settings,annual production of fathead minnow of up to368 kg ha)1 is regularly attainable (Dobie et al.1948). Factors driving differences in minnow densityamong regions has not been assessed, but abundanceof predatory fish, lake productivity and probability ofwinterkill are all probably important.

Zooplankton, macroinvertebrates (especially chir-onomids and amphipods) and detritus have all beenfound to be important components of fathead minnowdiets at various times of the year (Held & Peterka1974; Price et al. 1991; Duffy 1998). Similarly, wefound that all size classes of minnows consumed eachof these food categories in their diet at some pointduring the study. The exception was that we foundrotifers were also important at times, especially forlarval and juvenile fish. In addition to ubiquitousconsumption of detritus, larval fish consumed rotifersand non-rotifer zooplankton, juvenile fish consumed alarge percentage of zooplankton and macroinverte-

Larval fish

Log(number of larval fish·m–3)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

1.6

0.0 0.5 1.0 1.5 2.0 2.5

0.0 0.5 1.0 1.5 2.0 2.5

0.0 0.5 1.0 1.5 2.0 2.5

0

20

40

60

80

100

Log(number of invertebrates·trap–1)

0

20

40

60

80

100

Juvenile fish

Log(number of juvenile fish·m–2)

)teid ni sutir ted fo no itr oporp ( enisc rA

0

20

40

60

80

100


0

20

40

60

80

100

Adult fish

Log(number of adult fish·m–2)

0

20

40

60

80

100


0

20

40

60

80

100

Four superimposed data points

Fig. 6. Relationships between fish densityand consumption of detritus (left panels),and abundance of aquatic invertebrates andconsumption of detritus (right panels) bylarval, juvenile and adult minnows. Datapoints represent three estimates for each ofthree wetlands in two separate years.

Herwig & Zimmer

290

brates, and adult fish focused primarily on macroin-vertebrates.The predation pressure exerted by fathead minnow

for invertebrate resources shared with waterfowl andother wetland-dependent species suggests potential forcompetition (Duffy 1998), and reduced invertebratefood resources could have negative consequences forduckling growth and survival (Cox et al. 1998). Duffy(1998) estimated that consumption of aquatic inverte-brates by adult fathead minnow populations rangedfrom 333 to 1104 kg ha)1 year)1 in South Dakotawetlands. Our estimates ranged from 74 to2596 kg ha)1 throughout the growing season in 2001and from 584 to 699 kg ha)1 in 2002. Our maximumestimates were higher than those of Duffy (1998) in2001, but fell within his reported range in 2002.Additionally, aquatic invertebrate production rates inprairie wetlands have been estimated to range from 100to 500 kg ha)1 year)1 (Duffy 1998, based on literaturevalues) and from 1139 to 1836 kg ha)1 year)1

(Zimmer et al. 2001b, based on allometric models).Thus, estimates of consumption rates of aquaticinvertebrates by fathead minnow populations in this

study and Duffy (1998) approximate production ratesof invertebrates in these systems. This potentiallystrong predatory influence of fathead minnow popula-tions is consistent with several recent studies in the PPRdocumenting strong negative influences of fatheadminnow on invertebrate populations (Hanson & Riggs1995; Zimmer et al. 2000, 2001b, 2003).

Detritus was an important food resource for all sizesof fish. Averaged across wetland years, it accountedfor 40–79% of diet mass for larval, juvenile and adultfish. Larval fish accounted for 66% of the total mass ofconsumed detritus, despite the higher selectivity fordetritus by adult fish and the higher biomass of adultfish. Price et al. (1991) also found that detritus wasimportant in the diet of fathead minnow, with 54% offish sampled containing detritus in their gut. Priceet al. (1991) neither reported differences in thefrequency of occurrence of detritus among fish lifestages or sex, nor did they identify temporal patterns indetritus consumption.

If a choice between animal and detrital food existsand costs of capturing animals are not high, foragingmodels would predict that the fathead minnow shouldselect animal prey (high protein and energy) overdetritus to maximise foraging efficiency (Ahlgren1990). We found support for this prediction in juvenileand adult minnows, as both ages of fish showed anegative relationship between consumption of detritusand abundance of aquatic invertebrates. We alsodetected a positive relationship between consumptionof detritus and abundance of adult minnows. Takentogether, these results suggest that juvenile and adultminnows choose invertebrate prey over detritus wheninvertebrate prey are available and intra-specificcompetition is low, consistent with optimal foragingtheory (Schoener 1971). The absence of these rela-tionships for larval fish was somewhat surprising.Perhaps the increased energetic needs associated withbreeding and nest defence for maturing males (Unger1983) and egg production in maturing females (Grant& Tonn 2002) increase their selectivity for the higherquality invertebrate prey relative to larval fish. Orperhaps the cost/benefit ratio for consuming inverte-brate prey is simply larger for larval fish relative tojuvenile and adult fish.

The quality of detritus as a food source depends on itslevel of bacterial conditioning, but it is typically lowerin energy and protein relative to invertebrate prey(Bowen et al. 1995). Detritus has historically beenconsidered a food supplement that allows the fatheadminnow to survive when more nutritional food sourcesare low (Held & Peterka 1974; Price et al. 1991).Reduced quality of detritus, although, is presumablycompensated to some degree by higher overall availab-ility. More recently, it has been suggested that con-sumption of high-availability, poor-quality detritus, in

Adult fish

Arcsine (proportion of detritus in diet)

0

1

2

3

Juvenile fish

0

2

4

6

8

Larval fish

0 20 40 60 80

0 20 40 60 80

0 20 40 60 80

0

3

6

9

12

15

18

Sea

sona

l pro

duct

ion:

bio

mas

s

Fig. 7. Relationships between proportion of detritus in diets andproduction: biomass ratios in larval, juvenile and adult fish. Datapoints represent seasonal means for three wetlands in two separateyears.


291

conjunction with availability of high-quality inverte-brates may help sustain rapid growth and high survivalin fathead minnow populations (Bowen et al. 1995;Lemke & Bowen 1998). Our data support thishypothesis in that inclusion of detritus in the diet hasno negative effect on production. Despite consumptionof detritus (as percentage of total diet) ranging from 0%to 78% for larval fish, 0–76% for juvenile fish and23–94% for adult fish, we detected no relationshipsbetween seasonal P:B ratios and consumption ofdetritus for any age class of fish. Taken together, Lemke&Bowen’s (1998) laboratory results, and our field data,indicate that detritus is an important food resourcecapable of supporting metabolic processes and growthin fathead minnow populations. Consumption of detri-tus may be especially important for maintaining theextremely high minnow densities observed in PPRwetlands, as it weakens the link between abundance ofinvertebrate prey and minnow populations (Polis &Strong 1996; Schaus & Vanni 2000; Schindler &Scheuerell 2002). The ability to consume detritusresults in consistently high densities ofminnows despitefluctuations in the abundance of invertebrate prey.

The greatest source of uncertainty in this study isour imprecise estimates of fish density. The large errorbars in Fig. 2 would lead to error propagation andlarge error bars in Figs 3–5, and introduce error intothe regressions in Figs 6 and 7. Our estimates ofpopulation density are supported by the fact that ourestimates fall within the ranges estimated for fatheadminnow populations in other studies (Duffy 1998;Zimmer et al. 2001a), but imprecision undoubtedlyinfluences estimates and analyses in Figs 3–7. Asecond source of uncertainty is our use of bioenerget-ics parameters developed for larval perch to modellarval fathead minnow populations. These two speciesof fish are in different orders, and have vastly differentnatural histories. However, allometric patterns of fishmetabolism are strong, and application of adultbioenergetics parameters to larval fish is inappropriate(Post 1990). Given the strong pattern of metabolicallometry, we feel the bioenergetics parameters devel-oped for larval perch are likely a good approximationfor fathead minnow larvae. Overall, however, impre-cise estimates of fish density, coupled with our use oflarval perch parameters, reduce the precision of ouroverall estimates. Thus, our estimates and resultsshould be considered approximate attempts to assessthe role of larval fish and detritus in the populationecology of the fathead minnow.

At intermediate nutrient concentrations, shallowwater ecosystems are thought to exist in one of twoalternative states, either a clear water, macrophyte-dominated state, or a turbid water, phytoplankton-dominated state (Scheffer et al. 1993; Scheffer 1998).Recent work has indicated that this model also applies

to PPR wetlands, as these ecosystems exhibit adichotomy of dominance by either phytoplankton ormacrophytes (Zimmer et al. 2003; Herwig et al. 2004).The stable-state model suggests that fish stabilise theturbid state through two main mechanisms: resuspen-sion of sediment, and the suppression of zooplanktonpopulations. While appreciable resuspension of sedi-ment by fathead minnow feeding is unlikely, our resultssuggest that the high levels of zooplankton consump-tion exhibited by minnow populations could function tostabilise the turbid water state. Additionally, our resultsalso suggest fathead minnow populations may stabilisethe turbid-state in PPR wetlands through bottom-upinfluences on phytoplankton, where excreted nutrientsoriginating from wetland detritus contributes to high,stable algal densities (Braband et al. 1990; Schindleret al. 1993; Zimmer et al. 2006).

Overall, the ability to consume and utilise detritusas a food resource likely increases ecological influ-ences of the fathead minnow on wetland ecosystemsrelative to other fish species. Detritus helps sustainhigh densities of minnows and contributes to asubstantial flux of nutrients and energy from wetlandsediments into the water column. High density ofminnows and strong benthic–pelagic coupling, inturn, has potential to influence nutrient fluxes, watertransparency and abundance of invertebrates, phyto-plankton and submerged macrophytes in wetlandecosystems.

Acknowledgements

This project was funded by the Minnesota Department ofNatural Resources and NSF grants DEB 9815597 and 9977047to R. Sterner. K.D. Zimmer acknowledges financial supportfrom the Ecology, Evolution and Behavior Department at theUniversity of Minnesota. We would like to thank the followingindividuals for field assistance: L. Laurich, C. Augustin, W.Adams Jr, T. Herwig, A. Potthoff, M. Ward, M. Rud, C. Melaas,L. Rucinski. Special thanks to Vaughn Snook for the manyhours of field and laboratory assistance he provided. We thankDrs Mark Hanson and David Willis for their input on samplingprotocols and study design, and Brad Parsons and Jeff Reed fortheir input on study design and for providing field assistance.

We thank the staffs of the Fergus Falls, and Morris WetlandManagement District Office (USFWS) for the use of wetlandson WPA property, and continued support of our research efforts.Finally, we thank Dr Walter Duffy for his helpful advice onfathead minnow population and bioenergetics modellingapproaches. Daniel Isermann, Jeff Reed, Mark Hanson, JerryYounk, Don Pereira and Paul J. Wingate provided helpfulreviews of this manuscript.

References

Ahlgren, M.O. 1990. Diet selection and the contribution ofdetritus to the diet of the juvenile white sucker (Catostomus

Herwig & Zimmer

292

commersoni). Canadian Journal of Fisheries and AquaticSciences 47: 41–48.

Bence, J.R. 1995. Analysis of short time series: correcting forautocorrelation. Ecology 76: 628–639.

Bowen, S.H., Lutz, E.V. & Ahlgren, M.O. 1995. Dietaryprotein and energy as determinants of food quality: trophicstrategies compared. Ecology 76: 899–907.

Braband, A., Faafeng, B.A. & Nilssen, J.P.M. 1990. Relativeimportance of phosphorus supply to phytoplankton produc-tion: fish excretion versus external loading. Canadian Journalof Fisheries and Aquatic Sciences 47: 364–372.

Cox, R.R., Jr., Hanson, M.A., Roy, C.C., Euliss, N.H., Jr.,Johnson, D.H. & Butler, M.G. 1998. Mallard ducklinggrowth and survival in relation to aquatic invertebrates.Journal of Wildlife Management 62: 124–133.

Cummins, K.W. & Wuychek, J.C. 1971. Caloric equivalents forinvestigations in ecological energetics. Mitteilungen Inter-nationale Vereiningung fuer Theoretische und AngewandteLimnologie 18: 1–158 pp.

Cushing,D.H. 1983.Are fish larvae too dilute to effect the densityof their food organisms? Plankton Research 5: 847–854.

Dewey, M.R., Holland-Bartels, L.E. & Zigler, S.J. 1989.Comparison of fish catches with buoyant pop nets and seinesin vegetated and nonvegetated habitats. North AmericanJournal of Fisheries Management 9: 249–253.

Dobie, J., Meehean, O.L. & Washburn, G.N. 1948. Propa-gation of minnows and other bait species. Circular 12.Washington, D.C., USA: U.S. Fish and Wildlife Service.113 pp.

Dobie, J., Meehean, O.L., Snieszko, S.F. & Washburn, G.N.1956. Raising bait fishes. Circular 35. Washington, D.C.,USA: U.S. Fish and Wildlife Service. 123 pp.

Driver, E.A., Sugden, L.G. & Kovach, R.J. 1974. Calorific,chemical and physical values of potential duck foods.Freshwater Biology 4: 281–292.

Duffy, W.G. 1998. Population dynamics, production, and preyconsumption of fathead minnows (Pimephales promelas) inprairie wetlands: a bioenergetics approach. Canadian Journalof Fisheries and Aquatic Science 54: 15–27.

Euliss, N.H. Jr, Wrubleski, D.A. & Mushet, D.M. 1999.Wetlands of the prairie pothole region: invertebrate speciescomposition, ecology, and management. In: Batzer, D.P.,Rader, R.P. & Wissinger, S.A. eds. Invertebrates in fresh-water wetlands of North America: ecology and management.New York: John Wiley, pp. 471–514.

Golden Software. 1997. Surfer for Windows, version 6. Golden,Colorado, USA: Golden Software.

Grant, S.C.H. & Tonn, W.M. 2002. Effects of nutrientenrichment on recruitment of age-0 fathead minnows(Pimephales promelas): potential impacts of environmentalchange on the Boreal Plains. Canadian Journal of Fisheriesand Aquatic Sciences 59: 579–767.

Hanson, M.A. & Riggs, M.R. 1995. Potential effects of fishpredation on wetland invertebrates: a comparison ofwetlands with and without fathead minnows. Wetlands15: 167–175.

Hanson, P.C., Johnson, T.B., Schindler, D.E. & Kitchell, J.F.1997. Fish bioenergetics 3.0. Madison, WI: UW Sea GrantInstitute.

Held, J.W. & Peterka, J.J. 1974. Age, growth, and food habitsof the fathead minnow, Pimephales promelas, in North

Dakota saline lakes. Transactions of the American FisheriesSociety 103: 743–757.

Herwig, B.R., Hanson, M.A., Reed, J.R., Parsons, B.G.,Potthoff, A.J., Ward, M.C., Zimmer, K.D., Butler, M.G.,Willis, D.W. & Snook, V.A. 2004. Walleye stocking as a toolto suppress fathead minnows and improve habitat quality insemipermanent and permanent wetlands in the prairie potholeregion of Minnesota. St Paul, MN: Division of Fish andWildlife Special Publication Number 159, 42 pp.

Hewett, S.W. & Johnson, B.L. 1987. A generalized bioenergeticsmodel of fish growth for microcomputers. Madison, WI: UWSea Grant, Technical Report No. WISSG-87-245, 47 pp.

Klinger, S.A., Magnuson, J.J. & Gallepp, G.W. 1982. Survivalmechanisms of the central mudminnow (Umbra limi), fatheadminnow (Pimephales promelas), and brook stickleback(Culaea inconstans) for low oxygen in winter. EnvironmentalBiology of Fishes 7: 113–120.

Lemke, M.L. & Bowen, S.H. 1998. The nutritional value oforganic detrital aggregate in the diet of fathead minnows.Freshwater Biology 39: 447–453.

Markus, H.C. 1934. Life history of the blackhead minnow(Pimephales promelas). Copeia 1934: 116–122.

McCauley, E. 1984. The estimation of the abundance andbiomass of zooplankton in samples. In: Downing, J. &Rigler, F., eds. Methods of estimating secondary productionin freshwaters, 2nd edn. Oxford, England: BlackwellScientific. pp. 228–265.

Murkin, H.R., Abbott, P.G. & Kadlec, J.A. 1983. A comparisonof activity traps and sweep nets for sampling nektonicinvertebrates in wetlands. Freshwater Invertebrate Biology 2:99–106.

Payer, R.D. & Scalet, C.G. 1978. Population and productionestimates of fathead minnows in a South Dakota prairiewetland. The Progressive Fish-Culturist 40: 63–66.

Penczak, T. 1985. Phosphorus, nitrogen, and carbon cycling byfish populations in two small lowland rivers in Poland.Hydrobiologia 120: 159–165.

Peterka, J.J. 1989. Fishes of northern prairie wetlands. In: Vander Valk, A., (ed). Northern Prairie Wetland. Ames, IA: IowaState University Press, pp. 302–315.

Polis, G.A. & Strong, D.R. 1996. Food web complexity andcommunity dynamics. American Naturalist 147: 813–846.

Post, J.R. 1990. Metabolic allometry of larval and juvenileyellow perch (Perca flavescens): in situ estimates andbioenergetic models. Canadian Journal of Fisheries andAquatic Sciences 47: 554–560.

Price, C.J., Tonn, W.M. & Paszkowski, C.A. 1991. Intraspecificpatterns of resource use by fathead minnows in a small boreallake. Canadian Journal of Zoology 69: 2109–2115.

Schaus, M.H. & Vanni, M.J. 2000. Effects of gizzard shad onphytoplankton and nutrient dynamics: role of sedimentfeeding and fish size. Ecology 81: 1701–1719.

Scheffer, M. 1998. Ecology of Shallow Lakes. London, UK:Chapman & Hall.

Scheffer, M., Hosper, S.H., Meijer, M.L., Moss, B. & Jeppesen,E. 1993. Alternative equilibria in shallow lakes. Trends inEcology and Evolution 8: 275–279.

Schindler, D.E. & Scheuerell, M.D. 2002. Habitat coupling inlake ecosystems. Oikos 98: 177–189.

Schindler, D.W., Clark, A.S. & Gray, J.R. 1971. Seasonalcalorific values of freshwater zooplankton, as determined


293

with a Phillipson bomb calorimeter modified for smallsamples. Journal of Fisheries Research Board of Canada28: 559–564.

Schindler, D.E., Kitchell, J.F., He, X., Carpenter, S.R., Hodg-son, J.R. & Cottingham, K.L. 1993. Food web structure andphosphorus cycling in lakes. Transactions of the AmericanFisheries Society 122: 756–772.

Schoener, T.W. 1971. Theory of feeding strategies. AnnualReview of Ecology and Systematics 2: 369–405.

Smock, L. 1980. Relationship between body size and biomassof aquatic insects. Freshwater Biology 10: 375–383.

Spencer, C.N. & King, D.L. 1984. Role of fish in regulation ofplant and animal communities in eutrophic ponds. CanadianJournal of Fisheries and Aquatic Sciences 41: 1851–1855.

Stewart, R.E. & Kantrud, H.A. 1971. Classification of naturalponds and lakes in the glaciated prairie region. Washington,DC: Bureau of Sport Fisheries and Wildlife ResourcesPublication Number 92.

Unger, L.M. 1983. Nest defense by deceit in the fatheadminnow, Pimephales promelas. Behavioral Ecology andSociobiology 13: 125–130.

Vijverberg, J. & Frank, T.H. 1976. The chemical compositionand energy contents of copepods and cladocerans in relationto their size. Freshwater Biology 6: 333–345.

Zimmer, K.D., Hanson, M.A. & Butler, M.G. 2000. Factorsinfluencing invertebrate communities in prairie wetlands: amultivariate approach. Canadian Journal of Fisheries andAquatic Sciences 57: 76–85.

Zimmer, K.D., Hanson, M.A., Butler, M.G. & Duffy, W.G.2001a. Influences of fathead minnows and aquaticmacrophytes on nutrient partitioning and ecosystem structurein two prairie wetlands. Archiv fur Hydrobiologie 15: 411–433.

Zimmer, K.D., Hanson, M.A., Butler, M.G. & Duffy, W.G.2001b. Size distributions of aquatic invertebrates in twoprairie wetlands, with and without fish, with implicationsfor community production. Freshwater Biology 46: 1373–1386.

Zimmer, K.D., Hanson, M.A. & Butler, M.G. 2003. Relation-ships among nutrients, phytoplankton, macrophytes, and fishin prairie wetlands. Canadian Journal of Fisheries andAquatic Sciences 60: 721–730.

Zimmer, K.D., Herwig, B.R. & Laurich, L.M. 2006. Nutrientexcretion by fish in wetland ecosystems and its potential tosupport algal production. Limnology & Oceanography 51:197–207.

Herwig & Zimmer

294

Population ecology and prey consumption by fathead minnows in

Documents

Transcript of Population ecology and prey consumption by fathead minnows in