Arch.Hydrobiol. 158 1 57–74 Stuttgart,August 2003 Feeding patterns of migratory and non-migratory fourth instar larvae of twocoexisting Chaoborus species in an acidic and metal contaminated lake: Importance of prey ingestion rate in predicting metal bioaccumulation Marie-Noële Croteau 1, 2,LandisHare 1 * and Pierre Marcoux 1 With 6figuresand 4tables Abstract: Westudied diel variations in the feeding habits and migratory behaviors of two coexisting Chaoborus speciesin an acidicand metalcontaminated lake (Lake Tur- cotte, QC, Canada). Wefound that although the zooplankton community wasdomi- nated by rotifers, both Chaoborus speciesfed mostly on chironomids and crustaceans despite the relatively low abundance of theseprey types in the lake plankton. Chaobo- rus americanus larvaefed on those of Chaoborus punctipennis ,but not vice versa. The non-migratory species( C.americanus )fed throughout the day and night whereasthe migratory species( C.punctipennis )fed only atnight while in the watercolumn. The larger-bodied C.americanus consumed moreprey and had amorediverse diet than did the smaller-bodied C.punctipennis .Differencesin feeding habits between the Chaobo- rus speciesinhabiting Lake Turcotte (prey biomass, prey types) likely explain in part their ability to coexist. Attempts to predict Cd in the Chaoborus speciesusing our measurementsof Cd in their prey and their prey ingestion ratesmet with mixed suc- cess;although wecorrectly predicted higher Cd concentrations for C.americanus lar- vae than for C.punctipennis larvae, weunder-predicted absolute Cd concentrations. Wesuggest that studies such asours that arebased on analysesof gut contents of lar- vaecollected atintervals of 4hor longer likely underestimate prey ingestion rates. Key words: Bioaccumulation, diel vertical migration, feeding habits, coexistence, gut contents, cadmium. 1 Authors’address: Institut National de laRecherche Scientifique –Eau, Terre& Environnement (INRS-ETE), Universitédu Québec. C.P.7500, Sainte-Foy, Québec, G1V 4C7, Canada. 2 Present adress:U.S. Geological Survey, 345 Middlefield Road, MS465, Menlo Park, California, 94025, U.S.A. *Corresponding author; E-mail: [email protected] DOI:10.1127 /0003-9136/2003/0158-0057 0003-9136/03/0158-0057$4.50 ã 2003E. Schweizerbart’sche Verlagsbuchhandlung, D-70176Stuttgart 58 Marie-Noële Croteau, Landis Hare and PierreMarcoux Introduction Because larvaeof the phantom midge Chaoborus (Insecta, Diptera)occur over alarge range of chemicalconditions ( Hare & Tessier 1996, 1998), are widely distributed ( Borkent 1981), and areable to accumulate and tolerate high concentrations of tracemetals without ill effect( Croteau etal. 2002 a), they have been proposed foruse as biomonitors of cadmium (Cd) in lakes (Hare & Tessier 1996, 1998). To this end, models have been developed tore- lateCd concentrations in Chaoborus to those in lakewater( Hare & Tessier 1996, 1998, Croteau et al. 1998). However, such relationships areindirect because larvae of Chaoborus take up Cdfromtheir zooplanktonic prey rather than fromwater ( Munger & Hare 1997, Munger etal. 1999). Information on the feeding habits of Chaoborus could, therefore, be used to improve these models. Acase in point is Lake Turcotte (Québec, Canada), where model pre- dictions forCd in C.punctipennis did not explain measured Cdconcentrations forthis species ( Croteau etal. 1998). Furthermore, Cd concentrations in the two Chaoborus species inhabiting this lake ( C.americanus and C.punctipen- nis)differmarkedly ( Croteau et al. 1998). Although metalbioaccumulation can be influenced by physical (e.g., temperature: Croteau etal. 2002 b), chemical(e.g., tracemetal speciation: Campbell 1995, Hare & Tessier 1996) and physiological factors (e.g., metalassimilation efficiency: Croteau et al. 2001), diet-related variables such as the type, quantity, quality and Cd content of food arelikely central to explaining differences between sympatric species (Reinfelder et al. 1998). Majordifferences in diet have been reported for coexisting Chaoborus species (Sardella & Carter 1983, Hare & Carter 1987),although there areno published studies comparing the feeding habits of sympatric C.americanus and C.punctipennis larvae. Differencesin feeding habits could also help to explain the coexistence of these species ( Fedorenko 1975, Carter & Kwik 1977, Sardella & Carter 1983, Hare & Carter 1987),as could differences in their depths in the watercolumn ( Tsalkitzis et al. 1993) and the extent of their vertical migrations ( Carter & Kwik 1977, Hare & Carter 1987). To compare the diel feeding habits and vertical distributions of sympatric C. americanus and C. punctipennis larvae, wecollected benthic Chaoborus and zooplankton (including Chaoborus )fromvarious depths at4 hour intervals in Lake Turcotte. Tocompare Chaoborus feeding regimes, weexamined larval gut contents to estimateingestion rates, feeding periodicities and types of prey consumed. Lastly, weincorporated our measurements of prey ingestion rates and prey Cd into amechanistic Cd bioaccumulation model ( Croteau et al. 2001, 2002 b) to determine ifwe could predict Cd concentrations in the two Chaoborus species. Feeding of Chaoborus species 59 Methods Study site Wecollected zooplankton on September 19–20 2000 (stations Aand B)aswell as benthic Chaoborus on September 19–20 (station A)and September 23–24 (station B) 2000 from Lake Turcotte (48 ƒ 18¢ N, 79ƒ 04¢ W), Québec, Canada. Stations Aand B werelocated atthe samedepth (5m)on opposite sides of the centralbasin (50mapart) so asto encompassany horizontal heterogeneity in zooplankton composition ( Pinel- Alloul 1995, Folt & Burns 1999). Sampleswere collected at4hintervals over 24h, with sampling atstations Aand Bending and beginning, respectively, on the hour. This small(5 ha), shallow (5mmaximumdepth), highly acidic(pH 5) and fishless lake has been strongly influenced by nearby metalsmelters in Rouyn-Noranda, QC. Cro- teau etal. (1998) reported that the dissolved concentrations of severaltrace metals in this lake arevery high, e.g., 20nM for Cdand 5000nM for Zn, compared with concen- trations of thesemetals in pristine lakes(0.2 and 8nM for Cd and Zn, respectively; Hare & Tessier 1998). Atthe timeof sampling, the fall overturn of Lake Turcotte had already occurred sincetemperature, oxygen and pH werefairly constant with depth, –1 i.e., 13.9 ± 0.2 ƒC, 8.5 ± 0.1 mg O2 l and 4.9 ± 0.1, respectively (means ± S.D. of measurementsat 1mintervals on 19 September). Sample collection Wecollected planktonic Chaoborus larvaeand their potential zooplanktonic prey at 1200h, 1600h, 2000h, 2400h, 0400h and 0800h (timesof sunset and sunrise were 0645h and 1905h, respectively). Sampleswere collected using a0.05 m 3 Plexiglas plankton trap equipped with a64- mmmesh-aperture net. Collections weremade at 0.5mintervals from the surfaceto the bottom atboth stations Aand B.Sampleswere placed in 250 mLjars, to which chloroform wasadded, to anaesthetizeanimals so that they would not regurgitate their gut contents, followed by 10%formalin for preserva- tion. Wealso collected benthic Chaoborus larvaeat every sampling timeby taking a single sampleat eachstation using a15 ´15 ´ 15-cmEkman grab. Larvaewere isolated from the sediment by sieving through a0.5-mm mesh-aperture net then placed in a1-L jar, anaesthetized inchloroform and preserved in 10%formalin. For Cdmeasurements,we collected large numbers of zooplankton atmid-day on September 22 by hauling a64- mmmesh-aperture net atlake centre;the net’s path en- sured that plankton wascollected from all depths. This plankton wasplaced in plastic bags with lakewater.In the laboratory, weprepared two subsamples of this bulk plank- ton using a64- mmmesh-aperture nylon sieve. Subsamples wereplaced on piecesof pre-weighed acid-washed Teflon sheeting that werefrozen until Cd analysis. Wedid not measurezooplankton Cd over severalweeks prior to collecting Chaoborus (so as to integrate possible temporal changes in prey Cd) becauseunpublished data (M.-N. Croteau) show that Cdconcentrations in calanoid copepods aswell as in bulk plankton vary little over the 2weeksthat it takes Chaoborus larvaeto achievea steady statein their Cdconcentrations in the early spring ( Croteau etal. 2001, 2002b). 60 Marie-Noële Croteau, Landis Hare and PierreMarcoux Chaoborus crop contents Inthe laboratory, all Chaoborus collected weresorted to species( Saether 1972) and final (fourth) instar larvaewere differentiated by head capsule length, asgiven in La- row & Marzolf (1970) for C.punctipennis and in Fedorenko & Swift (1972) for C. americanus.All fourth instar larvaeof each Chaoborus specieswere dissected to re- move their crop ( Swift & Fedorenko 1973) and larval bodies and crops, opened to re- veal their contents, wereindividually mounted on microscope slides. Food itemsin crops wereidentified and prey densities in crops wereconverted to biomass based on the dry weights of various taxa given in Yan & Strus (1980) and Malley etal. (1989). Zooplankton communityanalysis All crustaceansand chironomids werecounted in plankton samples.Cladocerans and chironomids wereidentified to either genus or family levels. Copepods collected at Station Awereassigned toone of four groups: calanoid adults, cyclopoid adults, cope- podids or nauplii, whereasonly the adult stageswere enumerated in samplescollected atStation B.Rotifers wereidentified from samplescollected atStation Atothe genus level and counted in 1/16subsamples obtained using aFolsom plankton splitter. Al- though somealgae (mainly diatoms) werepresent