Aquatic Ecology (2006) 40:237 –245 Ó Springer 2006 DOI 10.1007/s10452-005-9028-9 Feeding on micro-algae in the invasive Ponto-Caspian amphipod Dikerogammarus villosus (Sowinsky, 1894) Dirk Platvoet1,*, Jaimie T.A. Dick2, Nellie Konijnendijk1 and Gerard van der Velde3 1Zoological Museum of the University of Amsterdam, Mauritskade 57, 1092 AD Amsterdam, The Netherlands; 2School of Biological and Food Sciences, Queen’s University Belfast, Medical and Biological Centre, 97 Lisburn Road, Belfast, BT9 7BL Northern Ireland, UK; 3Department of Animal Ecology and Ecophysiology, Institute for Wetland and Water Research, Radboud University Nijmegen, Toernooiveld 1, ED Nijmegen, 6525 The Netherlands; *Author for correspondence (e-mail: [email protected]; phone: +31-(0)20-525-6288) Received 18 February 2005; accepted in revised form 21 November 2005 Key words: Amphipod, Functional Feeding Group, Micro-algal feeding, Monoraphidium griffithii Abstract Feeding on micro-algae is shown in the invasive Ponto-Caspian amphipod Dikerogammarus villosus. Compared with controls, males, females and juveniles of this species significantly reduced the concentration in suspension of unicellular micro-algae. Juveniles had higher concentrations of algae in the cardiac gut than adults. The presence of these algae in the mid- and hindgut was also recorded. This feeding behaviour was filmed and the mechanisms involved are described and discussed. We comment on the use of the Functional Feeding Group (FFG) concept to classify feeding in amphipods. The role of being a feeding- generalist in aiding the invasion process is also discussed. Introduction conditions (Bruijs et al. 2001), we observed the following feeding mechanisms: carnivory, detritus Most members of the amphipod family Gam- feeding, grazing and coprophagy. Thus far, parti- maridae have long been considered, in regard to cle feeding on micro-algae has not been described their feeding habits, primarily as shredder of for this species. While there is evidence that am- detritus, with some recognition of omnivory phipods ingest micro-fauna and micro-flora (MacNeil et al. 1997). Although feeding on detri- (Hargrave 1970; Gowing and Wishner 1992; tus is important, there is a growing number of Gladyshev et al. 1999, 2000; Sushchik et al. 2003), observations of feeding on other types of food this has rarely been demonstrated (Blinn and (Dick and Platvoet 2000), which has lead to a Johnson 1982). Schram (1986) mentions several reconsideration of the position of these amphipods filter feeding methods in non-gammarid amphi- in food webs (Kelly et al. 2002a,b, 2003). In the pods in which antennae, mouthparts, and pereio- Ponto-Caspian amphipod Dikerogammarus villo- pods are involved. Gruner (1993) describe a sus (Sowinsky, 1894), which is invasive in western ‘passive’ filtering method in which particles are Europe (Dick and Platvoet 2000; Bij de Vaate collected by setose antennae (Haploops sp.), et al. 2002) and is able to survive ballast-water and an ‘active’ filtering method, in which the 238 amphipod’s ventilation current is used in combi- basic biology and the reasons for it being a suc- nation with setose pereiopods (Leptocheirus sp.). cessful invader. In this laboratory-based study, we This ventilation current is the result of an almost focussed on determination of the rate of clearance constant beating action of the three pairs of of a known concentration of micro-algae and the pleopods (Figure 1) as described by Dahl (1977). resulting effect on gut contents. Several possible Crouau (1989) mentions a method of filter feeding mechanisms involved in the ingestion of micro- for Mysidacea in which involvement of pereio- algae are described. pods, maxillulae and maxillae is suggested. Holmquist (1982) describes grooming for talitroids and summarizes those activities of the animals that Materials and methods are negatively influenced by lack of grooming: (1) respiration (epizoites hinder gas exchange); (2) Specimens of D. villosus and water were collected locomotion (damage to the pleopodal setation in- in May 2003 from the Gouwzee, a part of the flicted by epizoites diminishes locomotary abili- Markermeer (The Netherlands). The shallow lake ties); (3) sensation (chemo- and mechanosensibility has a soft bottom consisting of shell debris, coarse is affected by clogging up of pores and hindering sand, and clay. The shores are protected against movement of the sensillae); (4) water uptake in wave erosion by basalt rocks, which form the terrestrial species. In Talitroides sp. all the material primary habitat for D. villosus. The animals were ‘gleaned from the body is eventually handled by sampled with hand nets and immediately trans- the mouthparts’, after which the animals either ferred to the laboratory where they were kept in reject the material or ingest it (Holmquist, 1989). well-aerated containers of 65Â90Â15 cm with Particle feeding is common in organisms that 5 cm of lake water at a temperature of 19 °C. inhabit tubes, nests and other domiciles (Anderson Substrate of the containers consisted of debris 1994). D. villosus is primarily found among zebra from the sampling site. The light/dark regime was mussels or in crevices, where it keeps its antennae synchronized with the prevailing day and night stretched out, thus showing a form of domicolous rhythm in the field (17 h:7 h light:dark). behaviour. Determination of feeding on micro-algae in the family Gammaridae is important for recognising Algae feeding experiment their true ‘functional’ role in food webs. In addi- tion, because the present species, D. villosus,is Three size-groups of animals were used: 14 adult highly invasive, we require understanding of its males (length, 15 –20 mm), 14 adult females Figure 1. Outline sketch of D. villosus showing the direction of the water current that results from the constant beating action of the three pairs of pleopods. 239 (length, 10 –12 mm), and 14 juveniles (length, 3 – Photography and filming 5 mm). Single individuals were put in separate plastic containers (8Â8Â15 cm) with 70 ml of fil- With a Scanning Electron Microscope (JEOL JSM tered water from the lake (filter paper: pore range 35C) the microstructure of relevant body parts was 7 –12 lm), plus a control group of 14 containers recorded. The specimens were fixed in 3% glutar- without amphipods, all at 19 °C. The chlorophyte aldehyde in 0.1 M sodium cacodylate buffer and micro-alga, Monoraphidium griffithii (Berkeley), dehydrated in an ethanol series, critical point dried ) was added to a concentration of 5Â107 cells ml 1 and covered with gold. in each of the 56 replicates. These unicellular algae Five males of D. villosus were filmed with a are 30 –150 lm long, and are abundant in the digital video camera (JVC GR-DVL100) in order plankton and littoral zone of lakes (Streble and to determine feeding mechanisms. For this, a Krauter 2002) and in the Gouwzee (own obser- cuvette consisting of two glass plates separated by vations). No additional algae were added during a plastic hose was used, containing 30 ml of water the seven days of the experiment. Strong aeration with micro-algae in a concentration as in the above of the containers prevented settling of the algae. feeding experiment. Evaporation of water was compensated for by adding demineralised water (to avoid salinity changes). After seven days, the concentration of Results alga cells was estimated in each replicate using a 2 haematocytometer (0.0625 mm /0.2 mm depth). Alga feeding experiment Mean number of algal cells remaining was exam- ined in a One-Factor ANOVA with respect to The mean number of micro-alga cells remaining in males, females, juveniles and control groups, with containers among the four groups (F3,43=10.3, pair-wise comparisons of means with the Fisher p<0.001) differed significantly, with all three Protected Least Significant Difference Test. D. villosus groups significantly reducing the algae remaining as compared with the control group Gut contents (p<0.01 in all cases; Figure 2). No significant differences between the replicates within each Once the feeding experiment was completed, the amphipod group were found. gut content of 5 males, 5 females and 5 juveniles was determined by extracting samples from the cardiac stomach, mid- and hindgut from each Gut contents animal. The samples were separately squashed between a slide and a cover glass. For the cardiac D. villosus clearly consumed the algae (Figures 3 stomach, the number of algal cells ml)1 was esti- and 4). There was a significant difference among mated by averaging counts of 5, randomly chosen, males, females and juveniles in mean numbers of full frame microscope views per sample (diameter: intact algae of cardiac stomach sample 0.173 mm; sample depth: 0.013 mm; volume per (F2,12=3.51, p<0.05; Figure 3), with significantly ) frame 3Â10 4 mm3). The depth of the samples was more concentrated algae in juvenile than male calculated for a known quantity of water p<0.03) and female (p<0.05) stomachs (Fig- (1.5 mg=1.5 mm3 at 20 °C) and the resulting ure 3). The mid- and hindgut of specimens con- surface area under a cover glass (112 mm2): 1.5/ tained only fragments of algae, a strong indication 112=0,013 mm. The concentration was expressed of digestion. in number of alga cells ml)1. Means were examined in a One-factor ANOVA with respect to males, females and juveniles, Photography and filming with FPLSD post-hoc tests. Since the other part of the gut contained only fractures of the alga cells, The SEM pictures revealed various setal struc- only presence or absence of these fractures was tures. The long setae on the second antennae and scored. on the propodi of gnathopods 1 and 2 have the 240 25 (x 20 -1 15 10 5 number of remaining algae ml 0 Male Female Juvenile Control Figure 2.